Andrei Linde

Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.

During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.

We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.

Please contact [email protected] with any feedback.

Interviewed by
David Zierler
Interview date
Video conference
Usage Information and Disclaimer
Disclaimer text

This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.

This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.

Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.

Preferred citation

In footnotes or endnotes please cite AIP interviews like this:

Interview of Andrei Linde by David Zierler on January 26, 2021,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,

For multiple citations, "AIP" is the preferred abbreviation for the location.


Interview with Andrei Linde, Harald Trap Friis Professor of Physics at Stanford. This discussion continues from the interview with Linde conducted by Alan Lightman in October 1987. He provides a detailed history on his improvement of Alan Guth’s work on inflation, which Linde dubbed “new inflation” and subsequently “chaotic inflation.” Linde describes the impact of Perestroika on Soviet scientists, and the pressures he felt in preparing for a series of talks in Italy, which contributed to his development of “eternal inflation.” He discusses his formative early communication with American physicists including Lenny Susskind and Norman Coleman, he describes his two-year visit at CERN as the Cold War was winding down, and he explains his decision to accept a faculty appointment at Stanford. Linde describes the alternating feelings of hope and despair in the 1980s regarding the possibility that inflation could be observationally verified. He explains the intellectual origins of self-generating fractals that sprout other inflationary universes and the value of compactification theory, and he explains the cultural relevance of his Russian heritage which compels him to value theoretical notions and not treat them in a throwaway manner that capitalism can encourage. Linde explains how and why multiverses can be testable and he reflects on the obvious philosophical or even spiritual implications of this proposition. He discusses the impact of the discovery of the accelerating universe and dark energy and how WMAP strained the theoretical viability of inflation. Linde explains why many string theorists have moved into investigating theories of quantum information, and at the end of the interview, he reflects on the value of competing theories to inflation and why, ultimately, he wants to see a major convergence of theories so that the origins of the universe are well understood. 



Okay. This is David Zierler, oral historian for the American Institute of Physics. It is January 26, 2021. I am delighted to be here with Professor Andrei Linde. Andrei, it’s great to see you. Thank you so much for joining me.


Thank you for your interest.


Okay. So, to start, I would just like to note editorially to our researchers who will be listening to and reading this transcript we’re considering this interview with Andrei to be essentially part two, picking up of the last interview that Andrei did with Alan Lightman in October of 1987. So, we will pick up essentially from there and bring the narrative right up to the present.

So, Andrei, my first question is let’s just set the stage, if you can remember, in the mid to late 1980s. This interview with Alan that you gave in Cambridge, physically still, you were still in Moscow at this point. You had not moved over to the United States.


Yeah, that’s right.


What was your affiliation? What were you doing in Moscow at that point?


I was a senior scientific researcher in the Lebedev Physical Institute, which is similar to a professor, but we did not have any teaching responsibilities that we have right now.


What were you doing in Cambridge in 1987? Were you on a visiting professorship? Were you there for a conference?


No, I was invited there to give a series of Morris Loeb lectures, at Harvard.


Were you thinking about making a big move to the United States at this point, or that came later on?




Let’s start first with the state of play in terms of where things were with cosmic inflation circa 1987. It had already gone through various iterations since Alan Guth, of course. What was the state of the field by that point? What were the things that you were working on? How was the field refining from those original ideas earlier in the decade?


Well, I do not remember exactly what was in my interview with Lightman, so I may repeat something. In ’80, ’81 there was a paper by Alan Guth, and then everybody tried to improve it, but it did not work. Then I invented new inflation; that’s what I called it. It was somewhere in early summer of ’81, but I submitted it for publication only in October ’81 because of the difficulties to get a permission for publication, in Russia. Usually, it took at least 3 months to do it.

In 13-15 October, 1981, there was a conference on Quantum Gravity in Moscow. Lots of good people from abroad attended it, including Stephen Hawking. I gave a talk there on new inflation. Many people got excited about it and even suggested me to smuggle it abroad, but at that time I finally received a permission for publication. Then, next day after my talk, I translated a talk by Stephen Hawking at the Sternberg Astronomical Institute in Moscow University. In the beginning he was explaining to everyone the problems of the original Guth scenario. Then, in the middle of his talk, he unexpectedly said that I suggested an interesting solution of this problem, and I happily translated. And then he said that my scenario was wrong. I was a translator, so for half an hour I explained to everyone why my idea doesn’t work. Then I finished by saying that I translated, but I disagree. We started debating it with Stephen for about two hours. He invited me to his hotel, where we continued the discussion. Then he started showing me photographs of his family; we became friends.

Before I go any further, I should say that in parallel to that, there was also an inflationary scenario proposed by Alexei Starobinsky. It was invented in 1980, half a year before old inflation by Guth. Its goal was not to solve specific cosmological problems addressed by Guth, such as flatness, homogeneity, and horizon problems. Starobinsky attempted to solve the cosmological singularity problem. It did not quite work for that, but to be fair, the original scenario by Guth also did not solve flatness, homogeneity, and horizon problems. Meanwhile in 1981, Mukhanov and Chibisov have shown that expansion of the universe in the Starobinsky model produces density perturbations which can solve the problem of galaxy formation. But back in 1981-1982 the Starobinsky model did not attract much attention because it was based on modifications of the Einstein theory by quantum corrections, and it was not quite clear how to solve the problem of initial conditions in that scenario. It took a long time before cosmologists began considering his scenario seriously, and now it is very popular.

Returning to new inflation: After my paper and the conference in Moscow, things developed pretty quickly. In Summer 1982, Hawking organized a conference on inflation in Cambridge. It was the best conference that I ever attended. But during this conference it was understood that new inflation which I invented actually did not quite work. It produced too large perturbations of density.

Once it became clear, I tried to modify it, and I finally succeeded to do it a year later, in 1983, with invention of chaotic inflation. There’s a huge number of inflationary models now, but from the qualitative point of view, most of inflationary models right now are based on chaotic inflation scenario, which is different from the original setting of old inflation by Alan Guth and of my new inflation scenario.

In old and new inflation, the very early universe was supposed to be hot like in the old Big Bang theory, and it was supposed to be very large. How large? Well, at the very beginning of its existence, the universe was supposed to consist of billions of different parts, which did not know about each other, but mysteriously had approximately the same energy density. This means that old and new inflation did not really solve the homogeneity and horizon problems of the Big Bang.

This setting was so traditional and so strongly imprinted in minds of everyone that nobody wanted to get rid of it without a really good reason, and the reasons were there, but nobody wanted to look at them. But when we had seen that new inflation doesn’t quite work, so what could we do?

That is why I invented chaotic inflation, which was based on the simplest set of theories with the simplest potentials ever, like a potential of a harmonic oscillator, and just by magic this worked, there was inflation in the early universe. In this scenario one does not need the potential of a scalar field to have a flat top, as in new inflation. Instead of that one can consider a theory with a quadratic potential, or with a quartic potential, or many other potentials, and if they were not too steep, it also worked. So, there was a huge class of models where one can have inflation without any fine-tuning of the shape of the top of the potential. Inflation can start at the Planck density in the smallest Planck-size domain, and then this domain would inflate and become much greater that the part of the universe that we can see now. We don’t need the universe to consist of many, many different domains, one tiny speck of space in enough. And if the universe consisted of many different parts, then in some of them inflation may not start, and in some other it may begin. But the parts which do not inflate remain small and irrelevant, and the parts that inflate becomes large and uniform. One can create order out of chaos. That is why I called it chaotic inflation.

In ’83 I was in the US at the Shelter Island Conference, where we met again with Guth, Steinhardt, and Hawking, and I gave there a talk about chaotic inflation. Hawking immediately understood. Everything else that he was doing later in inflationary theory was in the context of this approach, to which he added his own changes and suggestions based on quantum cosmology. We were immediately in the same camp.

Meanwhile, most other people at first did not like chaotic inflation because the idea of initial thermal equilibrium was so deeply ingrained in our psyche that it was really difficult to get rid of it. Then a new generation of people came to inflationary cosmology, with the main desire to compare predictions of various models with observations. They did not care much about initial conditions and started associating chaotic inflation with the simplest model with a quadratic potential. [Chuckles] Meanwhile, there’s a huge family of models of chaotic inflation, including the models with plateau potentials, which became very popular lately. In my definition, any model where you do not need initial thermal equilibrium describes chaotic inflation. There is nothing chaotic about a quadratic potential. Chaotic inflation was about a possibility to create our universe even if it was very chaotic and inhomogeneous in the beginning. If inflation does not work, the universe collapses, or it is so ugly and inhomogeneous that we cannot live there. But if there is one small part of the universe, or one universe which just emerged in this particular state and inflated, then it becomes exponentially large and uniform, and can live there.

In parallel to that, in 1982 there was also an important development coming from Alex Vilenkin, who suggested that the universes can be quantum-created. At first, his expression for the probability of creation of the universe did not match the intuition of chaotic inflation, but then a year later I suggested how to improve his result, and shortly after that he came to a similar conclusion. Thus, now we have two different approaches pointing in the same direction.

And there was yet another line of thought, which has led to fairly unexpected conclusions. Back in 1982, in his talk at the conference in Cambridge, Steinhardt briefly noted that if inflation starts at the top of the potential in the new inflationary scenario, then the process of inflation always continues in some parts of the universe. During the conference, I have written a preprint suggesting that this process of inflation near the top may end differently in different parts of the universe, producing the multiverse consisting of different parts, which are so large that for all practical purposes they look like universes with different laws of physics in each of them. A very elegant approach generalizing this idea was developed by Vilenkin in 1983. But initially it worked only for the potentials with a flat top of the potential, like in new inflation, which was ruled out at that time.

And then in ’86 I found something that was completely unexpected for me. I found that in many models of chaotic inflation there are parts of the universe where inflation continues eternally, even if the potential does not have any top. The field wants to roll down to the minimum of the potential, but quantum fluctuations may bring it back again, and the process of inflation resumes. I called this process “eternal inflation.”

There are some people who love eternal inflation, like me or Vilenkin, there are some people who hate it—really, really seriously hate it for ideological reasons. In eternal chaotic inflation the universe becomes divided into many different parts, each of which develops according to its own logic. They all belong to the same universe, they are described by the same fundamental physics, but they can be very different. It is just like water: it can be liquid; it can be solid. The Atlantic Ocean contains lots of liquid water, but those who do not understand that water can be ice may crash into an iceberg.

So, that was the idea, that we should study all of this multitude of options and see what happens. It crystallized in my mind in ’86 after a set of unfortunate events. I don’t remember whether I told Alan Lightman about this in my interview in 1987. This was a rather unusual story. I don’t remember, really, so let me try to be on the safe side then…




I will tell this story because it was quite a shock for me. In ’85 I started writing a book on inflation, because everything seemed to be clear, and the theory seemed to be basically finished. The models were there. The theory was available, and while there was some difference of opinions whether the new inflation or chaotic inflation is right, it seems that we clearly understand what inflation is. Let me write a book explaining it. I started writing, and it was horrible. I hated it. I hated it because I like research when you discover new things, but when you already found something and are just telling about it for the fifth time, it is not exciting at all. Also, this was not the time when we had computers, we had just paper, pen, scissors, and glue. When you changed your mind about something, then you changed your mind about all these piles of paper. It was very depressing.

But on top of that, the government of the USSR at that time started Perestroika. Perestroika means they disrupted many things with good intentions. Gorbachev was trying to do a good thing, and one of the first things that he did was to abolish the system of getting permissions for scientific publications, because this was a redundant stumbling block. We were always three months behind everybody else in speed of publication, and, on top of that, we were often supposed to first publish our work in Russian. So, there were really a lot of silly things, okay?

They abolished the organization which gave permissions. But this implied that there was nobody who would give permissions and nobody who would tell that us that we can send a paper for publication, especially abroad. You could send it maybe to a Russian journal, but it would be translated to English only half a year later. We were just living like with our mouths shut for almost a year. Whatever you were doing, it was extremely difficult or almost impossible to publish it. That was really painful. Many years later, they returned to the status like it is now, that people just push the button, and the paper instantly travels to whatever place electronically, without any permissions, but in the 80’s a letter from Russia to the US would travel for about two months, okay? And before sending it, you would need to get lots of signatures allowing you to do it, so it was really weird.

In addition to that, we were gradually getting enough money to maybe buy a first very primitive car. I started learning for the first time how to drive a car, which was not easy at my age. When you are learning how to drive on the Moscow ice and your instructor tells you everything that he thinks about you in the language which you just barely know because it is not in any published dictionary, then it is very depressing. As a result, I was in a state which I don’t know what it was. Probably it was really a depression, or something similar, because I was lying in the bed, could not work, and I was reading detective stories. I was feeling ill, but all doctors told me that “You’re just fine.”

This was the Moscow winter, and then somewhere in maybe February of 1986, there was a call from my institute saying, “You must go to Italy.” I said, “What do you mean?” They said, “We have this Russian-Italian friendship collaboration, and you need to go to Rome, to Turin, and to Trieste and give popular lectures on astrophysics and astronomy to the citizens of Italy.” I was not very excited about it at all because at that time, there was a rule that we could travel abroad only once a year max, if they allowed it to you, and now they were calling me and saying, “You must go.” You know, if you have just once a year possibility, then you try to spend it for something which is really scientifically important, and not for giving popular lectures in astronomy. It was not what I was good at. It was just like giving lectures about something popular to somebody who doesn’t really care.

So, I told them, “I cannot go.” “Why?” “I feel awful. I am ill. I cannot go.” They told me, “Well, then please give us a certificate showing that you are ill.” Next morning, my wife Renata Kallosh (who is right now also a professor of physics here at Stanford), she went to the Lebedev Physical Institute, came to the Academician Ginzburg, who was the head of our theory division, and asked him to sign a letter saying that I was ill and could not go abroad. He was laughing, saying, “Ha, ha, ha. If he doesn’t want to go abroad, he can just say so and he won’t go abroad,” but he signed it anyway. She gave it to somebody there. A day later there was a call from there saying that “Well, today you are ill. Tomorrow you are healthy. If you just cannot go abroad at all, then tell us so,” and that was really, really scary because there were many people who would never go abroad for no reason, okay, or for some reason which they never knew.

To be in this category would be dangerous, so on Friday morning I took a taxi, which at that time was a financial decision for me, and I went to the hospital of the Academy of Sciences. Within one day, I made all necessary tests, and I passed all the necessary doctors. In normal life this would take about a week and a half, but I was really scared. I returned back home, and I was absolutely without any strength. I went to bed, and I was in that bed Saturday and Sunday. During that time, I was lying there and preparing this paperwork for me to go to Italy.

On Monday morning, I got up from the bed. I made another financial decision. I took another taxi to the Lebedev Institute. I paid our secretaries to instantly type all of this paperwork. I signed it in all possible places in the Lebedev Institute, which typically takes about a month, but I did it in one day because I was lucky that all people were present. So, I gave this signed paperwork to the person responsible for that.

A few days later, somebody called me and said, “Okay. We received your paperwork, everything is fine. You are going to Italy, but these people from Italy, they request that you send them a paper, like a text of what you’re going to give your talks about, so that citizen of Italy will be ready to understand what it is about.” I asked them, “When do you need it?” and they said, “Better tomorrow.”

I was feeling awful, but I understood that, on the other hand, if I did it, if I invented something in the next half an hour and I would be able to type it using my typewriter and insert equations and everything and tomorrow I would give it to them with everything, then it will go to Italy by diplomatic mail… For almost a year I was sitting with my mouth shut and now there’s full freedom, no permissions are required. I can just do it. Can I? So, I took my head in my hands like that and I started moving it back and forth. What can I invent in half an hour? What can I invent in half an hour so that I will type it and send it there?

In half an hour, I had this theory of eternal inflation with different parts of the universe and everything boiling out. You have all these universes with different properties appearing from one unique state, and it was amazing. And at that stage it became absolutely not important when it would be published or what! [Laughing] I sent them completely different stuff, just one of my old talks.

But a month later when I was going to Italy, I smuggled with me three papers which I had written during a month—two without any collaborators, one with two collaborators. I had all of these three papers in my luggage. First, they delivered my body to Rome and then I gave a lecture there. Then it was Turin, and I gave a lecture there, and it was about eternal inflation. People suffered probably listening to me. [Laughter]

Then they were delivering me to Trieste and some girl, from Russian-Italian friendship something, she wanted to show me Venice and I said, “Forget about Venice. I want to go to Trieste.” “Do you want me to show you anything in Trieste?” and I said, “Forget about it. I know everything in Trieste.” This was the International Centre of Theoretical Physics. Abdus Salam was there, and this was my first place abroad which I visited from Moscow, a dear place for me.

So, I came there and within two days I gave two talks there and I sent for publication three papers. Then I remember very clearly how it was in the evening of that day when I had done it. I took a train from Trieste to Rome to return later from Rome back to Russia, and I have never traveled since that time also, never traveled in such a luxurious train apartment, which was just one apartment for myself all with red wood and mirrors. [Laughing] I don’t know what it was. Crazy. I looked at the mirror and I saw an image of an old person whose body was slowly moving from Trieste to Rome. A month later, I was healthy. Everything was done and the papers were submitted. So that was how I invented eternal chaotic inflation, and that is what I was talking about at the Morris Loeb lectures at Harvard in 1987.


Oh, wow.


That was a fantastic time because I had seen so many wonderful people in US. That month, it was kind of a record-breaking stress for me. Within that one month, I gave lots of talks…Well, first I gave these Morris Loeb lectures. Including these lectures, I gave 23 talks in 11 universities in different cities traveling all around the US. I published three preprints there, and I think that one of these preprints was the reason why I am right now at Stanford.




It was about two globally interacting universes as a possible solution of the cosmological constant problem. I had this idea at Harvard. I published the paper there. I gave it to Sidney Coleman; he was skeptical. Then I arrived to Stanford and there was Tom Banks and Lenny Susskind and also Helen Quinn. I gave a talk there and then after the talk, we had a blackboard discussion. Immediately, in half an hour, Tom Banks and Lenny Susskind demolished my suggestion completely. There was nothing that remained from it. I was so unhappy, and it was so convincing what they said. Helen Quinn was standing there and translating what they said from their high-level theoretical physics language to my stupid understanding, so I was completely done. Then we went to the cafeteria, and on the way to the cafeteria, Lenny told me, “Yeah, but it seems that these effects which would contradict your suggestions, they are inversely proportional to the volume of the universe.”






Not a minor point!


And Banks told me, “Yeah, and I always wanted to invent something like that.” When I returned to Moscow, I received a paper by Sidney Coleman on a closely related issue. Then Tom Banks had a paper on the same subject, and Lenny Susskind had written a paper saying that both other papers are wrong. So, we were in the business of doing something interesting; it was called the “baby universe” theory. It caught imagination of many people for several years, and then it died by natural death.

We wanted to prove that vacuum energy must be zero, and this theory was able to prove it, kind of, under assumptions which were kind of plausible, but not exactly. We kind of proved that vacuum energy is zero, but we knew that our assumptions were vulnerable, so we were not saying that we had everything. Then eventually we found some real problems with this theory, got disappointed, and switched our attention to other things, but not everyone realized it and the wave of enthusiasm continued moving forward regardless. Now people do not want to remember this part of their life because, you know, if you are excited, you tell something and it does not quite go that way, then you may be annoyed when somebody asks: “How about your baby universe theory?” But nevertheless, there’s something in this set of ideas which still remains very interesting. It just did not work from the first go, but it was an interesting story.

Then in ’88, it was the first time when people from Russia were allowed to go abroad with their families. It was really, really unexpected. My wife Renata Kallosh, me, and two of my kids, we found ourselves in Geneva airport on the 31st of December 1988. We were moving through Geneva in our dark weatherproof jackets, and we were watching all these Swiss kids jumping from the buses all in different sweaters and coats of all colors, red, blue, yellow, pink, whatever. It was an interesting experience.

I remember how I went to a shop to buy watches for my children because they needed to go to school there. I went there, and I asked them…This is not quite related to, but it’s just how things were. I asked them, “How much are these?” and it was like 15 francs, maybe. For me it was lots of money. I asked them, “Do you have a warranty for them?” and they looked at me and said, “Are you Russian?”

A day later—well, maybe a couple of days later because it was New Year in ’89—I was at CERN, and I needed to get a photo for some document for an entry pass to CERN. There was a photo booth standing there covered by hundreds of happy people. I needed to go there, drop the 5-franc coin, and then it would produce a photo. So, I went there and dropped the 5-franc coin. I just had one in my pocket, and in a minute, the photo showed up there from the booth. I looked at it and I saw the person with an odd grim face like a convict, okay, and I understood that if I gave them this photo, they would never let me into CERN because I’m a dangerous person! [Laughs]

So, I thought, “Do I have another 5-franc coin?” Yeah, we found another 5-franc coin. This was a big decision. I looked carefully at all of the smiling faces on the booth. I dropped in my precious 5-franc coin. I went inside the booth, and I smiled and two minutes later a photo came from the booth. I looked at it and saw the photo of a smiling idiot because it did not come naturally, you know? I understood that they would not want such person to work at CERN. So, I gave them the first grim convict photo.

A week later, I came to the same booth. I needed another photograph for some other document. I dropped 5 francs. I smiled naturally and the photo was right. So, there are some things which you just cannot fake, or at least me. That’s why I’m not a movie star. (Probably not the only reason.)

We spent almost two years at CERN. At first, we were sure that we were going to return back to Moscow in a year, but CERN suggested me to stay there as a permanent staff member. Then somebody offered my wife Renata Kallosh a permanent position at the University of Annecy nearby. Then some people came from the University of Minnesota offering both of us positions there, and so suddenly things started changing in a way which we totally didn’t expect. Then we received an invitation from Fermilab, and from Lenny Susskind at Stanford.

This was not an easy decision absolutely because I had my parents in Moscow, and we were in very good relations with them. My car was waiting for me in Moscow, all oiled and everything. But we ended up at Stanford, maybe in part because we really liked people there. Well, maybe Lenny thought that it would be interesting to have us around. While at Stanford, we collaborated with him only once, on baby universes, which we at that time already didn’t quite like, but we decided to check what are the implications of related ideas for conservation or non-conservation of global symmetries. Right now, it’s one of the often-cited papers on the subject.




But even though we did not collaborate with him later on, his very presence is very important—he is such a powerful intellect.




He is so close in spirit to what I like in a physicist. I have a deep sympathy and respect for him, and just his presence at Stanford would be sufficient to make this place special. He built a theory group, person by person, of similarly spirited people, so I like the place very much.


Andrei, before we get more into your tenure at Stanford, I’d like to go back a little—a few questions on the collaborative side and the cultural side in the period before Perestroika. I’d first like to ask with new inflation, were you aware of what Andreas Albrecht and Paul Steinhardt were doing before their paper came out? Were you in contact with them?


I finished my paper on new inflation in early summer of ’81. I sent it for publication in October ’81 because of the delay with getting a permission. I was sure that the idea was so unusual that nobody would discover it, and I was not in a rush to publish. Sometimes in December of 1981, I received a letter from Paul Steinhardt—I have it here at Stanford—which said, in particular: “We were very excited to receive your recent preprint… upon seeing your results, felt we should put our present results in print.”

After this letter I waited to receive a paper from them shortly, but it was a rather long wait. Eventually, their short paper was sent for publication, three months after mine. I am not quite sure why they did not do it instantly. If one fully understands the importance of his work, then one can do what I did at the Lebedev Institute. One can just write a short paper, pay the secretaries to type it instantly and send it for publication in few days. In the US there is no need to get a permission, which took so much time in Russia. But somehow it took them three months. Their paper has some minor differences as compared to mine, but it’s essentially the same story. I don’t know whether one can call an investigation independent if one finishes a paper three months after mine, while having my paper in his hands. In theoretical physics, three months is a very long time, especially in the situation where the main progress is not in complicated and laborious calculations, but in a new idea, which is basically very simple.


This is before, of course, the internet. There’s no email. There’s no instantaneous communication, especially with you being on the other side of the world.


Yeah, there was no communication with Steinhardt on this issue, until he sent me a letter discussing my paper which he received. Albrecht was his student at that time.


Andrei, in those early years, ’81 to ’83, was there a sense that at some point these theories would be experimentally or observationally verified?


The hope…I think that hope clearly was. The problem was that we did not know when and how. Once—I don’t remember exactly the year, but it was maybe in ’83 or ’84, Igor Novikov, a prominent cosmologist, a collaborator of Zeldovich for many years, came to the Lebedev Institute saying, “Your theory has already been ruled out.” “Why?” “Because for explaining galaxy formation, we need perturbations of density at the level 10-3. But we did not find any perturbations of microwave background radiation at this level, so since we did not find them, it is against your theory.” At that time, I did not take it too seriously, but on the other hand, a prominent scientist said that, so there was an element of concern. I knew that it was indeed true that we needed to get the perturbations of a certain amplitude.

Thus, already in the early 80’s there was a suspicion that inflationary theory was dead, due to experimental evidence, okay? However, we expected that perturbations should be about 10-3 because at that time there was no such thing as dark matter. When dark matter appeared in the game, much smaller density perturbations were required, and that is what happened. When these perturbations were finally discovered by COBE in the beginning of the ’90s, then this was a confirmation of inflation. Whether you can prove or disprove something experimentally, this means that it is testable. And the argument given by Novikov at that time seemed quite generic, applicable to any inflationary model, not only to new inflation. New inflation was essentially, at least for me, dead after ’82, okay?

At the conference in Cambridge in 1982, there were four different groups which kind of simultaneously got the same result about density perturbations, except I had heard and seen this result reported by Alexei Starobinsky at the conference in Tartu about two months earlier. I remember it very vividly because it was quite damaging for inflation and for me personally. You see, new inflation is my theory. I am happily reporting it. I am such a lucky guy, and then Starobinsky gives his talk and says, “But the amplitude of the perturbations is too large.”

Then I saw, about a month later, a paper by Hawking saying that everything is fine there. When we came to Cambridge in the UK, everyone was trying to understand the situation with the perturbations. There were four groups studying perturbations at the conference: Hawking, Starobinsky, Guth, and a group including Bardeen, Steinhardt, and Turner. In the beginning, all of these groups had different opinions. After some discussions, Hawking completely changed the main conclusion of his paper, and now I cannot find his original preprint. Starobinsky did not change his conclusion, which he reported in Tartu, before anybody else.

Then they gradually converged to similar conclusions and published their results, though the paper by Bardeen, Steinhardt, and Turner was published a year later. It was a huge paper, very difficult to read. I came to Moscow with this bunch of papers, except for this last one, but all the talks were available. I gave them to Mukhanov and Chibisov, who developed the theory of these perturbations in the Starobinsky model a year earlier, in 1981, and they could not make any sense out of these papers because they were complicated, and many arguments, different from one paper to another, were kind of intuitive.

Then Mukhanov started working on it, under my pressure, in particular because the results of this set of papers was more applicable to new inflation, during which the Hubble constant did not change much. It was not easy to apply them to chaotic inflation. Why would I need to know details of new inflation if I already knew that it did not work? That is why I asked Mukhanov to develop a theory applicable to other inflationary models including chaotic inflation, and he did. He is a really powerful person, but he completed it only in ’85. After that he developed a general approach to such perturbations, so he is really behind the rigorous approach to that.


Andrei, the criticism of chaotic inflation, given how significant it was and from how many directions it was coming from, what was the value in this debate? What were some of the common themes you were perceiving in the criticism to chaotic inflation, and how did that move the whole enterprise forward?


I don’t think that there was much of a criticism, except for…It’s just like nobody wanted it in the beginning except Hawking because new inflation and old inflation, they were based on such a conventional clear-cut idea that in the early universe there was a thermal equilibrium, and everyone believed it. Then the scalar field, because of this equilibrium, was at the top of the potential and everyone believed it. It was based on the theory of high-temperature phase transitions which was proposed by my thesis advisor, Kirzhnits, and then we developed it together with him. And the theory of supercooling, which is the basis of old inflation, was developed in our paper with Kirzhnits in 1976. So, in the beginning of the 80’s this part became the well-established feature of the hot Big Bang story.

But I knew that there was a subtlety about it. The origin of thermal equilibrium was not clear: elementary particles should collide many times to come to a state of thermal equilibrium, and this takes time. If you estimate how much time is needed, you will see that the time from the Big Bang to the beginning of inflation may be too short for that, especially if you want, as we wanted, to have a theory with very small coupling constants to avoid problems with large density perturbations. So, thermal equilibrium was questionable. Everybody knew about it before. Steven Weinberg had written about it in one of his papers long before all of this inflationary story, and he thought that maybe there were some quantum gravity effects at the moment when the universe was born. Maybe this was how the equilibrium was established, but there was no real mechanism. Then we understood that different parts of the universe, at the moment when it was born, did not talk to each other, and if they did not talk to each other, how about the thermal equilibrium, or any equilibrium, okay?

So, then one may start thinking about it and say, “Let us try to make a theory which would not require any thermal equilibrium, or any equilibrium at all.” We [had] something similar with Starobinsky, with respect to his own model, and he told me something like, “Oh, but who cares about the equilibrium?” or something like that. At some moment, I just decided, well, let me check what happens in the simplest models with quadratic or quartic potentials. Then I found that the universe may inflate while the field is still outside of any equilibrium at all, so we do not need the field to be at the top of the hill and to make this top of the hill very, very flat. For me, this was the crucial moment when I realized that inflation does not need any artificial constructions and fine-tuned potentials, it does not require any assumptions about the hot Big Bang, thermal equilibrium and supercooling. If everything works in a much, much simpler theory, why would one need anything more complicated?

But old ideas do not die easily. When I was writing my book, which was published in 1990, its substantial part was dedicated to explaining why there is a problem of initial conditions in new inflation and how it is solved in the context of chaotic inflation. But this is very difficult to explain.

I used to say that any three people discussing the theory of initial conditions will always have at least ten different opinions about it, in part because one should clearly define what initial means. Mathematicians say, “Okay, let me have a set of equations,” and they impose initial conditions. Hawking would say, “My initial conditions are that there are no initial conditions. Quantum cosmology takes care of everything.” I would say, well, I call “initial” the moment from which I can solve my classical equations, and that is at the Planck density because before that, when density was higher than the Planck density, close to the Big Bang singularity, quantum fluctuations were so strong, our normal concepts of space and time did not apply, and I do not know how to impose initial conditions on anything. But when density becomes much smaller than Planck density, this is not an initial point, it is already an intermediate evolution, which depends on initial conditions. So, it makes sense to impose initial conditions as soon as it becomes possible, i.e., close to the Planck density.

Then I realized that it is sufficient to impose initial conditions in the smallest possible domain of the Planck length, at the Planck time. And if one assumes that the universe was relatively homogeneous on the smallest possible scale at the earlies possible time, it is sufficient for inflation to begin in many simple versions of chaotic inflation.

For me, this was such an enormous relief, but it was unconventional, and many people, still now, have problems with that because even now most of the college textbooks teach students that the universe must be incredibly large from the very beginning. This was the dominant idea before the invention of inflation, and nobody noticed that this assumption is no longer required.

Later I learned—and there was a set of different people who studied this idea—that one can impose initial conditions in the toroidal universe which looks like an expanding bagel. For some special reason, it may be much easier to start inflation in a universe having the shape of a bagel than having a shape of a sphere. Then the problem of initial conditions can be solved even easier. And after inflation the universe becomes so large and flat that nobody living there can see that it is a torus. But if you tell anybody that our universe looks like a bagel, they will think that you are crazy.

So, this is a psychological and sociological issue. Back in 1988, Hawking in his book A Brief History of Time said: “In my personal opinion, the new inflationary model is now dead as a scientific theory, although a lot of people do not seem to have heard of its demise and are still writing papers as if it were viable.”


It’s a zombie theory, Andrei. It won’t die.


Unfortunately, what Hawking said in his book in 1988 is still valid now. When I open the textbook which I am using for teaching undergraduate students, it says that our universe was incredibly hot, then it cooled down, and there were cosmological phase transitions, and supercooling, and then inflation. This theory died almost 40 years ago, but it seems that nobody knows it. I am sending messages to the authors of these books, asking them to correct it. They reply, “Oh, yes. We will discuss it with our colleagues,” and that’s it. People who write these books, they do not know.


Andrei, I’d like to ask a question on the cultural and the international side of things. Of course, the story with inflation for you begins in a very dark period of US-Soviet relations. This is before Gorbachev. It’s before Glasnost and Perestroika, and as you indicated, these restrictions you felt personally as a scientist, this was a part of your reality. I’d like to ask specifically in your decision to come to Stanford in 1990, this is right in the middle of the process whereby the Berlin Wall falls, the East Bloc is falling apart, and the Soviet Union is rapidly disintegrating. In what way did the international situation in Eastern Europe influence your career decision-making at this point?


Well, first of all, some qualifications. Intellectual freedom was unlimited in the place where we were working at the Lebedev Physical Institute. Wonderful place. From the point of view of doing physics, just wonderful. The best scientists; huge institute. On the other hand, if I wanted to collaborate with anybody working abroad, or communicate, writing papers, getting signatures, it was intimidating.

We had Andrei Sakharov in our theory division, and we were trying to, well, kind of insulate him from the outside pressure, which was severe. Our theory division was under the microscope. An interesting property of our theory division was that when two people were talking and a third entered the room, they continued talking. It was nontrivial because who knows who might enter your room and what are you talking about. There was trust, friendship and mutual respect between our people.

My first travel abroad was in ’76. It was totally unexpected that I was able to travel to Italy to Abdus Salam in Trieste, because most of those who wanted to go abroad were asked to admit that Andrei Sakharov was anti-Sovetchik. This is a special term which essentially means that he is an enemy of the people. I was not a member of the Communist Party, but if anyone wanted to travel abroad, this should have been approved by the Party Committee of the Institute.

I remember how it was. About 20 people from different parts of the Institute and three of us from the theory division where there. They first invited all people not from the theory division and approved their travel without any problems. Then they invited the first of Anatoly Shabad and asked: “Do you agree that Sakharov is anti-Sovetchik?” I was listening because even though it was a thick oak door, you could hear what was happening inside. He said something which immediately strengthened my respect to him: “Yeah, before coming here, I read literature about that, and I’ve found that anti-Sovetchik is a certain well defined legalistic term. If a person is anti-Sovetchik, then he must be convicted in a court of justice, and imprisoned. Because Sakharov is not convicted, I cannot take on myself a responsibility claiming that he is anti-Sovetchik.” For half an hour they pressed him, and pressed him, and pressed him again, so that he would admit that Sakharov was anti-Sovetchik, and he did not. And they did not sign his documents for travel to Cuba.


To Cuba!


To Cuba.


Wow! That says it all!


Then the second of us came inside, and within a minute, not only he agreed to admit that Sahkarov was anti-Sovetchik, but he even volunteered to give a lecture on this at Lebedev Institute. That was a total shock for me.

Then I entered there, and I looked at the faces around and I thought, who is going to ask me this question. But what happened was totally unexpected for me.

Two hours earlier, I met some guy whom I barely knew. He was a nice person, and he asked me, “Aha. I know that you are going to be invited to this Party committee today.” I said, “Yes.” He said, “Do you know what the percentage of growth of national product of Moldavia this year?” I replied, “Are you insane?” He said, “No, no. I would suggest maybe you should know.” That was weird, but I went to the library. I got the newspaper Pravda. I read everything there very, very carefully.

So, what happened is that apparently this person was a member of this Party committee. When I went into that room and I looked at everyone wondering who was going to ask me the question about Sakharov, he asked me, “What is the percentage of growth of the gross national product of Moldavia this year?” I answered, and I guess nobody expected it. Then somebody else asked me, “And what is happening in Angola now?” and I was able to answer. They asked me more questions and I answered all of them, and they told me, “Okay, you may go.”

I went out. I was totally exhausted morally. They didn’t ask me this question! I closed my eyes, and I’m not a religious person, but I said, “Oh my God, thank you for saving me from this humiliation.” Then I opened my eyes and I found that I was standing under the portrait of Brezhnev. So, you remember these things, you know?


Yes. Yes. Yes.


Nevertheless, life scientifically was excellent, and when our family came to CERN many years later, on 31 December 1988, I was sure that I will return back to Russia, but… I did not. So, the fact that I did not return—these facts contributed, but maybe just when you go somewhere for a week, abroad for a month, then it is not realistic for you. When you go somewhere with a family for a year and then you are staying a little bit longer, maybe temporarily but just a little bit longer, then slowly you start understanding that, well, maybe it’s a possibility for you to stay here longer. This happened gradually, and of course all of this experience, and the experience related to Sakharov whom we visited in his exile in Gorky was emotionally not very pleasant. All of this, it gradually became part of us, so I don’t know what exactly triggered the decision, but many things together. I believe that the main reason was purely scientific. Life is very different when you can publish anything you want without any permissions, instantly communicate with your colleagues all over the world, and meet all of them at the conferences without any problems.


So, is this to say when you accepted the offer at Stanford, did you recognize you’d be making a life for yourself in the United States; you would not be going back to the Soviet Union (or Russia) thereafter?


Well, it was a tricky situation. For several years I still signed all of my papers like Stanford University and Lebedev Physical Institute. I was still officially a member of the Lebedev Institute, okay, and that was in part because we wanted to protect our friends at Lebedev because you know, if you fly away and leave Russia, you are not considered kindly, but also your friends who surrounded you could be held responsible one way or another. Then gradually, the number of people who were staying abroad had grown up a little bit, and it became kind of normal. Eventually, it became clear that we were going to stay here. But there was always a small probability of who knows what. My parents were in Russia. I needed to take care of them somehow, but nevertheless, our understanding when Renata and I accepted positions at Stanford was that we are here for good.


Your relationship with Lenny Susskind—I’m curious if, specifically, his expertise in string theory was interesting and valuable for your research.


Not this. You know, he has a profound intuition, and when you are speaking to him, you know he loves what he does, and he is open and sincere, and deep. Like his idea of string theory landscape—those words were coined by Lenny Susskind, but they appeared after in 2003. Kallosh, Kachru, Trivedi and I invented something which is called the KKLT scenario, a possible way to stabilize vacuum in string theory. Then Michael Douglas and his collaborators, found that there are enormously many different vacuum states in string theory stabilized by this mechanism. The question is still debated, but the possibility emerged. Then Lenny jumped and said what happens when you unify cosmology and KKLT, and he not only had written a preprint, but he had written a book about that. It was also influenced by our discussions at that time. I was giving talks there about the universe consisting of different domains, but he joined all of these pieces together in something very nice and clear. That made this scenario very popular among string theorists.

I recall how maybe a few months after that I visited KITP, the Institute in Santa Barbara. There was a conference there, and several people at the conference approached me and asked me, “What do you think about string theory landscape?” I told them, “I love it! This is a great idea,” and I started explaining why. I continued explaining it, and then one of them pointed a finger at me and said—and I do not want to give his name because it was actually pretty friendly, but he pointed at me and said, “And you are the worst!” [Laughter] What could I do? So, yeah, we influenced each other, even though there was no formal collaboration, but it was really necessary to be in the atmosphere where we could influence each other.


Andrei, I’d like to ask about the intellectual origins of this idea of the self-reproducing inflationary universe as you described in your 1994 Scientific American article. Were these ideas that mostly began with your time at Stanford? Is the idea of self-generating fractals that sprout other inflationary universes—does this really begin before your tenure at Stanford, or this is all happening in those early years in the 1990s?


It was much earlier. Well, first of all, in order to come with this idea, you need to have some pieces of background. Back in Russia, even before invention of inflation, I attended talks by Iosif Rozental on anthropic principle. He speculated that in the very early universe the laws of physics were not fixed yet, and we live in the universe with these parameters stabilized at the particular values allowed by the anthropic selection.

I was attending his talks out of politeness. He was a friend of my parents. It looked somewhat weird, but on the other hand, it was also interesting. Why weird? Because we knew that the electron mass is just a constant, taking the same values everywhere in the observable universe, and so it is not something you are going to choose. You can speculate what could be if it were different, but these speculations are empty if it is just a constant of nature. It seemed clear that the universe is not a country which you can choose, the universe is given to us, we can learn its properties, but it is the way it is, there is no choice.

Of course, there were other people with different views. There were books by Martin Rees about the anthropic principle, there was a large book about it by Barrow and Tipler, which I have seen at the bookshelf of Mike Turner, whom I visited at Fermilab. I saw this book and I said, “Oh, you have the book!” and he said, “Oh, you want it? Please take it.” I said, “No.” I was at this time still in Russia and I told him, “That’s okay. I will try to order it in Moscow.” He said, “No, please take it.” I told him, “Ah. Are you going to destroy my scientific reputation because you know that I like the anthropic principle and then I will follow it and then I will be just morally dead?” [Chuckles] He said, “No. I want to stop destroying my own reputation by the mere fact that I have this book on my bookshelf.”




So, that is how it was. But the process by which I came to this started in this conference at Cambridge in ’82. All of us were giving talks there. In his talk, Paul Steinhardt said that actually, there is something metaphysical about our universe, inflationary universe, because when inflation starts at the top of the hill, then sometimes it stops and sometimes it may continue. There was some description of this, but it was just a few sentences in a talk on a different subject.

I started thinking about it, and then I had an idea how one can describe this process and use it to solve the singularity problem. It’s like you have expanding space, and you produce one bubble and produce another bubble and you can continue producing bubbles, but space between the bubbles continues to expand, so in most of the space you are still standing here on the hill, and maybe this was your initial state, standing on the hill, but this standing on the hill state is a part of de Sitter space which is nonsingular. It has curvature given by Hubble squared and whatever, so it’s just a constant, and if it is a constant, then you start the universe, actually many different bubbles, many different universes and you don’t need initial singularity. So, I decided that this was a cool idea and maybe we’d just write something together with Steinhardt and we started discussing it.

But then my thought evolved in a completely different direction, and I started seeing that in addition to maybe a possibility to start a nonsingular universe, we have a possibility that a spontaneous symmetry breaking goes in one direction and in another direction and in a third direction, and that is anthropic. The universe creates exponentially large parts of the universe with different properties, with different dimension of spacetime, with different value of the electron mass. In some theories, e.g., in grand unified theories, there are many vacua. It was known already at that time. So, in some of these vacua, there is no difference between weak and strong interactions in the electromagnetic; all this was unified. This was one of the vacuum states. But the same theory allows vacuum state wish symmetry breaking SU(4) × U(1), and also with symmetry breaking SU(3) × SU(2) × U(1). So, even in new inflation, when you are falling from the top, you can end up in a universe with different properties.

I knew it because I studied this process, not just speculated about it. I was writing a paper about this process of generating different types of vacuum states. I was writing it in parallel with my paper on new inflation cosmology because I thought that nobody would believe my paper on new inflation because it was such an unusual story. So, I started developing first a theory of tunneling to prove that what I had seen there was a real thing, then studied how it works in realistic theories like SU(5), and I found that you can go in many different directions. So, then one has a real possibility of creation of huge bubbles, universes, with different laws of low-energy physics, and that is what is necessary for a scientific justification of the anthropic principle.

During the conference in Cambridge, I had written a paper which described two different ideas—first, that you might solve the singularity problem, and then that you also have this anthropic landscape, so to speak. I typed it at this conference in Cambridge on my own, and during the last few days before the conference was over, it was issued as the Cambridge University preprint. I distributed it among all the participants, none of whom remembers it right now. [Chuckles] Because this was still too weird at that time, okay?

Then on the flight back, Starobinsky and Novikov, who was there also, told me that the idea about absence of singularity was wrong because we knew that something similar like that could be possible also in the flat de Sitter space, but we know that the flat exponentially expanding de Sitter space is geodesically incomplete. A complete space is the closed de Sitter space, which first collapses and then expands, and if you are considering only the region of expansion, this is not the whole story.

I understood that this was right, that part of the idea which I had written a paper about actually does not work, but I already sent this paper to Physics Letters. I received a letter from them accepting my paper for publication. At that time, I had very stringent criteria with respect to myself: if a paper is only half right, it should not be published. So, I withdrew the paper from Physics Letters. It was a painful decision and probably a stupid one because this was the first paper describing what was later called inflationary multiverse.

However, we were writing proceedings for this conference, and for me, proceedings of the conference in Cambridge means that everybody would know about it, okay? So, who cares about Physics Letters? This is the Proceedings of this famous conference, the best in my life. I had written there a large description of inflationary theory and ended up with this discussion saying that it does not solve the singularity problem, but it allows you to make a scientific justification for the anthropic principle because it gives you an option in which universe to live.

At that time, I already knew that in string theory, and in the general theory of compactification, people considered a possibility that our universe is 10-dimensional, and we ended up in a 4-dimensional space-time because of some specific choice of compactification. But then the question was why our universe is three-dimensional. If six dimensions have compactified, then we get three special dimensions, but if only four compactify, then we get five. So, why is our universe three-dimensional? Then I remembered that Paul Ehrenfest in 1917, which was simultaneously the year of the Great October Revolution, and also the year when de Sitter invented his de Sitter solution…




So, in 1917, Ehrenfest explained why we live in the three-dimensional universe: because in the universe of higher dimensionality, planetary systems would be unstable and atomic systems would be unstable. In a lower dimensional universe, there is no gravitational attraction between isolated bodies at all. If you have two isolated objects, then in two- and one-dimensional universe, these two objects do not interact. In Newtonian theory, they interact quite strongly, but in general theory of relativity they don’t. That is why we can live only in the three-dimensional universe.

Of course, in 1917, we did not have string theory, the theory of compactification, and inflationary cosmology. But with the invention of inflation the possibility that different exponentially large parts of our world may have different number of dimensions becomes real, and this would explain why our part of the universe is three-dimensional.

I’ve written it in an appendix because it is kind of speculative. It was published, but this was just an appendix in the proceedings. Who reads such stuff?

Then in ’83 also I had written a paper on this multiverse appearing in the context of the supersymmetric SU(5) model. In supersymmetric SU(5) model, you have a dozen of minima of the potential with equal energy, i.e., a dozen of different vacua, okay? People realized at that time that if the universe originated in the hot Big Bang, then originally, at high temperature, the only minimum was the SU(5) symmetric minimum. When temperature drops out in this particular supersymmetric version of the SU(5) model, the potential acquires a dozen of minima of practically the same depth, but one cannot go from the SU(5) minimum to the minimum with physics of our type. It was proven by Steven Weinberg long ago. This suggests that this theory cannot describe our world.

But inflationary theory quantum fluctuations can occasionally bring from the SU(5) minimum (I will call it red) to the minimum where symmetry is SU(4) × U(1) (green), or to the minimum with our type of physics with symmetry SU(3) × U(2) × U(1) (blue).

Then you may start with a red universe, but after inflation it becomes divided into red, blue and green universes, and you live only in one place where you can live; that’s all. There are some places with SU(5) symmetry nonbroken. We cannot live there. Somebody else can live there, but we can live only in the blue universe. I gave a talk about it at the Shelter Island II conference in ’83. That was also when I gave a talk about chaotic inflation. There were some illustrations, which have shown the universe divided into different parts, with dragons eating you if you enter the Dragon universe (after Norbert Dragon, who found this particular minimum in the supersymmetric SU(5)).

So, it was fun, but it was not really fully intellectually satisfying for me because it was in the context of new inflation. The energy scale of inflation there was not very large. Therefore, it was not clear whether quantum fluctuations were powerful enough to bring the universe from one vacuum to another. And to change the dimensionality of space, you better have inflation at energy density close to the Planck density.

So, in ’86, when I found that this universe during chaotic inflation enters the stage of self-reproduction at density which may approach the Planck density, this concern was eliminated. In the end of my paper of ’86 describing it, I had written that in string theory, people complained that they have too many solutions, which they considered as a problem of string theory. But I argued that this is not a problem but a virtue of the theory, because inflation can produce so many different states with all possible compactifications and all possible dimensionalities, and the more such states, the better is the chance that we can live in one of them. Once I realized it, for me there was no way back. This was so amazing. But then again, there are some people who consider the abundance of choices as a defect, because historically many of us wanted to prove that the universe did not have any choice. This is like trying to prove that water can only be liquid, never solid.


Mm-hmm [yes], mm-hmm [yes].


And I think that…Maybe this is something which I mentioned in my interview in ’87, that it is very monotheistic. You know, many people want their religion to be the only one possible. There is one god, one rule, and if you do not believe this god, then he will punish you and your family and all of your relatives and whatever. So, something like that, something which unfortunately is deeply rooted in the psychology of many people, unfortunately in many countries, okay? So yeah. The possibility of having the universe consisting of different parts with no particular prescription in which universe you should live—it is maybe closer to Buddhism, or you can find something related in the Bhagavad Gita, but it is certainly not the idea which many people even now would easily accept.

So, there are some people who really deeply, sincerely hate it, and I understand that well. For me, a Russian, going to Safeway for quite a while was a challenge. Suppose you want to buy coffee. In Moscow, the last years when I was there, it was not easy to find coffee. Then somebody would call you and say that they were selling coffee right now in some particular shop. You did not ask whether it’s Arabica or Colombian. You just ran there and buy as much as you can because there were some limits. You bring it home and give it to your friends, and they do not ask you what kind of coffee it is. So, now you go to Safeway and suddenly you have an abundance of choices. You look at the shelves; there are too many types of coffee, you do not know which one to buy, you are desperate, and you go away not buying anything. So, I totally understand this desire to have one simple prescription.

But on the other hand, and maybe for the same reason, when you live in the country where we had one single prescription and everything else was considered wrong, it’s kind of refreshing having a universe where you have a democracy. You have a chance to live in the red universe or in the blue universe and the green universe, and then you do not care, as long as this universe is suitable for your type of life. If somebody wants to live in the red universe, let them, and I’m living in the blue universe and it’s nice there. In inflationary cosmology, I probably cannot travel to the red universe. Not because I will die on the border and not because somebody will shoot me, but because there is an enormously high potential energy wall separating these two universes. If I am young and stupid and I just will penetrate through this wall, then I die there because particles from which my body is made may not exist in another universe.

So, this is a different type of story, but it nevertheless allows you some freedom. Those people who have this freedom from childhood may be longing for a single rule. Some other people who when going to vote were always given a ballot with a single candidate in it may be longing for freedom of choice. This might explain why Vilenkin and I love anthropic considerations and multiverse, and some other people really sincerely hate it. So, what can I say?


Andrei, I wonder if, in thinking about multiverses and in your reference to God and belief in things like that, I wonder if inescapably with this pursuit you are bumping into non-scientific considerations because of the inherent un-testability of the multiverse as a proposition.


I don’t believe that it is untestable. Let me start with a negation, okay?


Please, please.


Okay. Here is the story, and this is how usually I end my lectures of multiverse. There is a great British philosopher, Sherlock Holmes, who said that “when you have eliminated the impossible, whatever remains, however improbable, must be the truth.” So if you try to explain some facts which people were trying to explain for decades and failed, and there is one paradigm which allows you to explain these facts, then you accept this paradigm because you don’t have any alternative.

We want to understand why the cosmological constant is so small. At first, we were trying to explain why it is exactly zero. Stephen Hawking wanted to explain why it is zero. Sidney Coleman wanted to explain why it is zero. I was trying to explain why it is zero. Tom Banks was trying to explain why it is zero. All of us were trying to explain why it is zero, for many, many years, and it did not work.

Simultaneously with trying to explain why it is zero, we were also trying to explain why it is maybe nonzero, but extremely small. One of these explanations related to quantum cosmology and fluxes was in an appendix of my review of inflation in ’84. Another one appeared almost simultaneously in the paper written by Andrei Sakharov when he was in exile. His paper and my paper appeared almost simultaneously—we were discussing these possibilities when I was visiting him in Gorky. He assumed that there are many dimensions, and if there are many dimensions, some of them may be compactified differently and you have many different options. Then in his paper he referred to my scenario of many different options of compactification. He argued that if one has exponentially many different compactifications, it provides us with exponentially many choices of the cosmological constant. We can live only in the subset of all possible universes where the cosmological constant is sufficiently small. The same ideas many years later resurfaced in the context of the string theory landscape, which relied on the multiplicity of choices of fluxes and compactification.

So now we have this possibility to solve the cosmological constant problem. This does not mean that we already did solve it, but this is a paradigm which allows us to address this issue. Then you have a problem: why a proton has the same mass as a neutron with a very large accuracy. One may say, well, if one would be heavier than another, then the heaviest will tend to decay to the lightest. For example, if a neutron is too much heavier, then there will be a universe consisting only of protons. There will be no chemical abundance of anything. Protons will only make hydrogen, no carbon or oxygen, okay? If protons were heavier, then they would decay and there would be only neutron stars. There will be no normal matter, okay? Either it’s only hydrogen, or neutron stars, so fine balance between proton and neutron masses is necessary for our existence, so it’s anthropic. But the normal reply is that this argument does not make any sense because our universe is given to us in one copy. You do not choose where you live, and so it may be life could adapt to that, okay? But it’s not the universe which could adapt for our life. They could not meet each other midway, okay, life and the universe. And now consider electron mass. It is about 2000 lighter than a proton. If it were heavier, or if it had a different charge, we would not be around. But this argument again does not seem to make any sense, if the mass of the electron is just a constant, a parameter of the fundamental theory.

So, there are many coincidences which anthropic principle could explain, but these explanations don’t make any sense, unless we are using the concept of multiverse. Then we can say, following Sherlock Holmes, that since we failed to provide any explanation of many mysterious coincidences, and we have only one paradigm which allows us to explain these coincidences, then this new paradigm must be the truth. In a sense, physicists play a role of a detective, and we also do not like mysterious coincidences. A good detective considers these coincidences as experimental data. In this sense, mysterious coincidences, requiring using the concept of the multiverse for their explanation, are the experimental data supporting the theory of the multiverse.

About a religious story—it was maybe not exactly religious. It may be philosophical, but yes, long before I invented new inflation, when I was in high school, I was trying to understand what is real, matter or mind. I was reading some science fiction books, and some of them contained ideas that forced me to study philosophy. And I found that some people writing about philosophy are very smart but do not necessarily know what they’re talking about, and some others are really deep and know what they’re talking about but cannot come to any final conclusion. What they all agreed about is that our knowledge comes from observations, i.e. from our consciousness, so what will it be if I make consciousness the basis of everything?

Pretty quickly—I was at that time a schoolboy—I invented the theory which seemed to explain all miracles of telepathic contact. This theory assumed that consciousness, the feeling of pain, or happiness, is a really existing object. It really exists. It’s just like, well, glass, apple, it is real.

Based on this idea, I invented a theory of telepathic contact, and I was very proud of it. Then a year later I realized that in my construction, I heavily used the concept of simultaneity. And in special theory of relativity the concept of simultaneity is a very tricky issue. My beautiful theory crumbled. I realized that I can talk about philosophy, and try to reinvent the way we think about our world, but I will be at risk of saying many stupid things if I do not know the rules of the game, if I do not know physics. This reinforced me in my desire to become a physicist, but simultaneously, I could not forget that sometimes when one thinks about the standard opinions of people which we often take for granted, there are sometimes some question marks about that. When some people, me included, say that they know answers, sometimes the question mark is still hanging around, okay?

Shortly before I invented new inflation, I was in contact with yoga people in Moscow, and it was very interesting because these people were really deeply philosophical, but also a little bit religious of a special kind. Eventually I just left this small involvement, but I’ve read books on Indian philosophy, and I found that, independently of religious undertones, it contains some powerful archetypes of thought. And later I found it very useful to be exposed to different cultures of thought, because it may help to withstand moral pressure coming from the culture of a country where one lives. And it is OK to ask questions which many people were afraid to ask.

When I am thinking about creation of the universe from nothing, from singularity or from whatever, the first question that comes to mind is—what happened? What was created first, the law which describes the universe, or the universe which contains the law, because if you say that there was a law of physics and the universe was born in accordance with the law, then the question is on which blackboard the law was written, okay? If there was no blackboard—what does it mean that the law existed? Do we think that law is something that can exist without matter? Seriously? Or if the universe existed before the law, does it mean that the universe was unlawful until some moment when the law gradually formed? How did it happen?

In the multiverse kind of story, you must push it high up to quantum cosmology and maybe beyond. There is a concept of “It from bit” by John Wheeler, okay? Or maybe the wave function of the universe consists of many branches, describing different universes with different laws of physics and different observers. The whole structure is extremely complicated, and our goal would be to find an unambiguous interpretation of this picture. We are still far from it.

I remember how I was discussing philosophy with Allan Sandage, a famous astrophysicist who was trying to measure the Hubble constant many, many years ago. He was philosophically minded, and it was nice to talk to him. It was really interesting. At some moment, we were discussing solipsism and how our knowledge of the universe grows, and he told me: “You know what? I was a solipsist in my youth, and then I met some guy in a restaurant and started talking with him. This person learned that I was a solipsist, and he told me, ‘Oh, what a coincidence! I am also a solipsist! So nice to see you! There are not so many of us!’”


[Laughs] That’s great!


So, you know, we should have some sense of humor discussing our theoretical constructions.


Andrei, I wonder if you can talk about the role of computers as you were continuing to think about multiverses, given the fact that simulations were so important to this research.


I am not a computer expert, but I did study tunneling using the awfully primitive computer at the Lebedev Physical Institute in the ’80s. I was studying tunneling there, and the computer was spitting on me all these punched cards. I was suffering, but then one of the results of this investigation was that sometimes you tunnel not to the minimum but to a point close to the top of the potential. Later on, this result became an important ingredient of new inflation.

I am not good with computers even now, but if I want to see something, I am playing with Mathematica. Sometimes I am learning how to use it from students, sometimes I do it myself. It is a very powerful tool, sometimes it really enhances your mind. I am very visual, okay, so I understand something better if I draw pictures. If I draw potentials using Mathematica, I can immediately understand how a field moves there. It’s helpful.


Andrei, what was the impact of the discovery of the accelerating universe? What was the impact of that on your research?


You mean the discovery of dark energy in 1998? Huh. First, it was a discovery of a very slow inflation which happens right now. In effect, it was a confirmation of the main idea of inflation that expansion of the universe may accelerate. But it had additional significance for the theory of inflation in the early universe. Most of inflationary theories predicted a flat universe, which means that the ratio of the density of the universe to the critical density called Ω must be very close to 1. But in the ’90s, people started bombarding us with questions, saying, “You predict a flat universe, but we know that the universe has Ω = 0.3.” That was coming from many different astrophysicists. I started looking seriously, and there were some papers by Neil Turok and collaborators saying that it’s possible to have a universe which is inflationary but open. It was interesting for me, and I studied their work. I found that for the simple models with simple potentials, which they described in their talks but not in their paper, their scenario just does not work. But in general, it may work and then you may have universes with Ω not equal to 1 coming from inflation.

I invented one model which really worked, and I was happy about that, but the circumstances in which I invented it will immediately tell you how shaky the scenario was. I invented it at a flamenco show in Barcelona. I was at a conference in Barcelona, but one of my sons insisted that while in Barcelona I must see at least one flamenco show. I went to a flamenco show, but I made a mistake. The show started three hours later than I arrived, but it was a long train ride coming to this place, so there was no way to come back.

I thought I must work, so what did I have with me? I had only a pen and a metro ticket—what I could write on the metro ticket. I wanted to invent a theory of open inflation that would work, and my problem was to make it so simple that it would fit on this metro ticket. I had nothing but a pen and this ticket. [Laughs] I drew the potential on the ticket which could work, okay? Then I came to these people in the show which had not yet started, and I asked them, “Well, I understand that the show is later, but can I sit here at the table and work?” Soon there was no free space on the ticket was just ending. I asked them for some paper, and they did not understand what it was that I was talking about—maybe toilet paper or whatever—but they brought me some paper. Then they brought me some wine and then the flamenco show started three hours later. By that time, I had a theory of an open universe, okay? [Laughs] The potential really fit in this metro ticket, but it was ugly. But it worked. But it was ugly. So, what can you do with an ugly theory? Well, it is better than nothing, you publish it, but you admit that it is ugly.

Then with two of my friends from Japan, we continued studying it. We found another example of an equally ugly theory which also worked, and they developed a theory of perturbations in that theory. It is possible to do it, but it was ugly. Ugly means that you must be really inventive, want to twist the arms of the theory, and if you really want to, it is possible to do so. So, having inflation with Ω < 1 is possible. It is also possible to have Ω > 1 in inflationary theory, but it is even more difficult than having Ω < 1.

Then Lenny, a really smart person, came with the anthropic justification of the possibility of why it is not so bad having an open inflationary universe. It was similar to the anthropic solution of the cosmological constant problem. Like with the tiny cosmological constant, the universe must be just a little bit open, okay? So, it was not a total disaster, but really to draw the potential which would work for Lenny’s scenario as well as for mine, you do not draw a nice potential; you draw something ugly which is custom made for explaining experimental data. Ugly does not necessarily mean wrong, but we are trying to avoid fine-tuning in constructing our theories as much as possible.

In ’98 I was at a conference on cosmology at UCLA, and there were these talks about the discovery of dark energy. With dark energy, suddenly we had an addition of vacuum energy to the total energy density, making Ω close to 1, just as inflation predicted. Well, at first their results seemed suspicious. I called my friends, astrophysicists. They told me, “No, we are not so sure,” but then next day they said, “Yeah, maybe it is true, in fact. Yes, maybe.”

Then I called my wife Renata Kallosh. She was at that time at KITP giving a talk, and I told her about that. The same day when I told her, she was talking with David Gross at KITP, and he told her, “Oh Renata, it is such a pity that inflationary theory does not work.” She asked, “Why do you say so?” and he said, “Oh, because I was at a conference at Princeton, and they were completely adamant that Ω = 0.3. I am so sorry that inflationary theory is ruled out,” and Renata said, “No. Andrei just called me and said that everything is fine because dark energy was discovered. That was at the conference at UCLA, so everything is fine.” He looked at her and said, “I was at a real conference, at Princeton, not at UCLA, and they explained that Ω = 0.3, and therefore inflationary theory does not work.” The next day, all the walls of KITP were covered by newspapers announcing the discovery of dark energy.

Even now, some people ask, “Is it possible to disprove inflation?” Inflation has passed several tests where it had kind of a near-death experience. One near-death experience was when Novikov came and said, “Your theory is ruled out because perturbations at the level 10-3 are not found.” Then we learned that dark matter saves inflation, and inflationary perturbations were discovered by COBE.

Then we were told that Ω = 0.3, and few years later inflation was saved because dark energy was discovered. You would not really argue that dark matter and dark energy conspired to help inflationary theory. These guys, dark matter and dark energy, they do not have brains to conspire, but it looks as if two important ingredients, which are not even fully understood yet, conspired for consistency of inflation.

We could be able to do something else, but this “something else” would make the theory much less attractive. It would not be ruled out unless a better theory explaining how much energy of the universe would be produced.

You know, one would not discard a theory that does not look nice until we find a theory which is even nicer, okay? But the level of happiness would be less, and I do not say that I am completely happy with what we have now. No, I am not, because we want to have a theory which is based on best versions of the fundamental theory, like string theory. It is possible, but very difficult. Supergravity—very difficult but doable. We have several different models of supergravity inflation which work perfectly well. Okay. So, we have very high goals, and we definitely will not reach all of them. We just work with the best we have.


Andrei, as the Planck satellite was getting ready to release data, did you know immediately how important this was going to be for inflationary cosmology?


Oh, we knew it weeks before that, months before that. This was yet another near-death experience for inflationary theory. [Chuckles] In 2012 first we heard rumors, very persistent rumors, that the WMAP satellite discovered something which indicated large non-Gaussianity of inflationary perturbations. All simplest versions of inflationary theory predict almost exactly Gaussian perturbations, but, according to the rumors, WMAP satellite tells us that it is not so. Gaussian is like when you have a coin, and it flips normally, as expected. But if the coin is bent, then some statistics of which side lands up becomes different from what’s expected.

There were rumors that in summer 2012 there would be a meeting in Germany, I believe, where all of these statements about non-Gaussianity would be revealed and then we would be ready to bury nearly all best and simplest models of inflation. It would not mean that inflation is wrong, but it would mean that you need to make it much more complicated. Many of us were preparing for the worst, but all students and post-docs were happy because they knew that they had a problem to solve, and this problem was difficult but solvable. Let’s do this or that or invent some other mechanism of producing non-Gaussianity. I was writing papers explaining that it is possible; it is possible to make models producing non-Gaussianity. It is difficult. It requires often making some ugly changes to the theory, but if it works, ugly—okay, I can live with ugly if this is the only option, okay? If nature tells us something, then our ideas of what is ugly may be changing. How about the top quark mass being almost 200 times heavier than a proton? It’s kind of ugly, but we are all happy that the top quark is there, because without it the theory would be inconsistent. So, ugly but consistent is okay. The Higgs mass is very heavy. Nobody expected it to be in the range it is right now, its value may even imply that we live in a metastable vacuum state. It is ugly. Nobody wants to live on an atomic bomb, okay, but the experimentally observed value of the Higgs mass may imply that we live in an unstable vacuum state which will eventually explode, destroying our universe. It’s scary ugly, but we still call the electroweak theory beautiful. So, what can I say? Non-Gaussianity would be okay, but it would be ugly. That is why we were waiting for Planck data on Gaussianity.

In early 2013, I was in Europe. I was standing in the line boarding the airplane to fly back to the US, and I received e-mail from Renata that Planck just had a press release with their data. The title of her message was: no non-Gaussianity. So, with all of these many, many different new data, these were the most important words in the title of her message, no non-Gaussianity. I was standing there waiting to board the airplane, really close to it, and I opened the iPhone which she lent me. I did not have my own iPhone at that time, but I have it now. I used this iPhone like a point through which I downloaded Planck papers to my notebook which I opened there, standing in line. I received several most important papers by Planck. I was reading them during the flight. I was almost crying looking at the beautiful figures. Wonderful. They were just stunning. And the next data releases, in 2015 and 2018, they were even better, but this first data release was crucial. For me, this was extremely important. There’s no non-Gaussianity.

But there was also something else. They, following WMAP, killed my simplest inflationary model with a quadratic potential, they did not find gravitational waves at the level which this model predicted. They have found instead that the best model to fit their data was the Starobinsky model invented in 1980, a few months before the old inflation by Guth. In addition, there was some model called Higgs inflation, which was first studied by some people in Canada and the US in the 80’s, but then this paper was not really known by many. Then Shaposhnikov and Bezrukov developed it many years later and pushed it forward. These two different models, the Starobinsky model and Higgs inflation, give the same prediction for the amplitude of gravitational waves, and also for the tilt of the spectrum of inflationary perturbations. Both of these predictions were in perfect agreement with Planck data, and my quadratic model was not.

Well, so when many people tell you that you are drunk, sometimes you need to think about it seriously, and we started trying to understand, me and Renata together. We were trying to understand what was so special about these two models, Starobinsky model and Higgs inflation.

We found that it is possible to take the Starobinsky model and reformulate as a theory of a scalar field, but at an intermediate stage of this reformulation you see a theory with a strange kinetic term, so kinetic energy looks unusual. If you do the same with Higgs inflation and do a similar transformation, then at an intermediate stage, you find a similar kinetic term.

Then we went to a conference on cosmology, again at KITP. Our son told us that we should stop at one particular burger place on the way because their burgers were really good. We stopped at this burger place, but there was a long line, so we were waiting a long time. We found an empty table and used napkins to write something while waiting. And then we formulated the theory which started in a crazy way. It described two different fields. One of them had positive potential and kinetic energy; another had negative potential and kinetic energy. But this second one did not really have any physical degrees of freedom and could be transformed away. But the fact that it was present in the original model made the original two-field theory very symmetric. You look at the theory, and it has a very large symmetry, and we love symmetry in physics. Interestingly, when you eliminate this extra field, you get a potential which is very similar to the potential of Higgs inflation—different, but very similar. Huh. And then we realized that our simple model could describe much more general potentials, but whatever potential we start with, for each of these potentials, we have the same cosmological predictions, similar to what was in the Starobinsky model and in Higgs inflation. So, we have a much more general class of models which have predictions providing a perfect fit to the Planck data—all of the predictions, a huge class of models which we discovered in this burger place, okay? [Chuckles]

Few days later, we reported it at the conference at KITP. We called these models cosmological attractors because one can substantially modify these theories, but their predictions remain unchanged. That is why the Starobinsky model and Higgs inflation have similar predictions: they belong to this general class of cosmological attractors.

Then later on, together with Ferrara, Porrati, and Roest we found another class of models, which is even broader, α-attractors. With α-attractors, we were able to modify predictions for the amplitude of gravitational perturbations, while keeping the same prediction for the slope of a spectrum, which was the main inflation-related quantity constrained by the Plank satellite observations. Thus, now we have a broad class of new theories which fit the Planck data, and which also have flexibility to describe possible finding or not finding of inflationary gravitational perturbations. If not for this pressure of experimental data from WMAP and especially from Planck, we would not invent these models.


Andrei, just to bring the narrative up to the present, what have you been working on in recent years? What has been your research post-Planck satellite?


Well, one of the problems is very theoretical: how about a theory of initial conditions for this class of models which have potentials of the type of Starobinsky, or α-attractors? In these models, inflation cannot start at the Planck density, which was an important advantage of the simplest chaotic inflation models studied in the 80’s.

So, there was a challenge. Can we solve the problem of initial conditions for this class of models? Fortunately, we have found many simple ways to do it. One of them was especially nice and general, but a lot of effort was required to prove that it actually works. This was one part which left me intellectually satisfied, but nevertheless, we continue working on it just to be totally sure about some details of it.

Another direction is developing good models—good models of inflation in the context of supergravity or string theory. We work on it together with Renata Kallosh and many of her collaborators. Lots of new interesting concepts have been invented, and I’m quite happy with the progress. But the progress is incomplete, in part because string theory is very complicated and not fully explored. This is a real problem, and as a result many string theorists moved into investigation of the theory of quantum information, or of the high temperature superconductivity. But we hope that cosmology may help us in some way.

I will give one example. When Planck issued his last data release in 2018, they gave a list of inflationary models which they considered to provide the best fit to the Planck data. Our alpha attractors are among the best, but there was also some equally successful set of models which they called KKLT inflation models, using the names of the authors of the string vacuum stabilization scenario invented in our paper with Kachru, Kallosh, and Trivedi of 2003. At first, I did not recognize this model. Soon after the KKLT paper on vacuum stabilization, its authors together with Juan Maldacena and Liam McAllister attempted to develop inflationary scenario based on KKLT. We almost invented everything, and it was very satisfying, but as I said, string theory is very complicated, and some of its elements are not firmly established. At some moment we decided to align our scenario with the intuition that was more in parallel to what string theorists in our group were thinking about at that time. And once we did it, the inflationary model became very cumbersome. The previous version of this scenario worked great, but it was not quite aligned with some conjectures which were popular at that time. Thus, we removed our original model from the main body of the paper, but decided to preserve it in the Appendix, just in case. So, that’s how it worked. [Laughs] We decided to keep it in the Appendix.

Then many people studied the model which we presented in the main body of the paper, and there was a general sense of dissatisfaction. It did not quite work, so people turned their attention to other models based on string theory. One of the most intellectually appealing models is the model proposed by Eva Silverstein and her collaborators. It is consistent with the Planck data, but it may not provide the best fit; it is still too early to tell.

So, we looked at these Planck papers, and Renata said, “Hmm. Interesting. They call it the KKLT model, and indeed, this is the original version of the string inflation model which we put in the appendix. This is what we preserved just in case. Maybe we can return to it and check whether one can have some kind of string theory or supergravity justification for it.” We started studying it together with one of our Japanese colleagues, Yusuke Yamada, and it looks like it has a reasonable interpretation in terms of string theory, or at least in string theory motivated versions of supergravity. We published a paper about it, and then we decided to take another look at theoretical consistency of all other models favored by Planck.

We ended up with a very short list which consists of alpha attractors (including the Starobinsky model and Higgs inflation), and the models of the type of our KKLT inflation model. Each of these models can describe a certain range of values of the so-called spectral index ns, but practically arbitrary values of the tensor to scalar ratio r, which corresponds to inflationary gravitational waves. At the Planck figure describing their data for ns and r, theoretical predictions look like a collection of vertical stripes. By finding experimentally in which stripe we are in, we could say which of these models has a better chance to be correct.

So, we are waiting for new data. We are working preparing ourselves for this eventuality. We often collaborate with Renata. She brings to it her knowledge of string theory and supergravity, and I am bringing my knowledge of cosmology, but there are some issues about cosmology that she already knows much better than I do. It works. It is complicated. It’s complicated because each of us may see the truth in a different way. It is a real collaboration where collaborators are very critical to each other, and that’s exactly what we want because we want to find what it true, rather than to find something which would make your collaborator happy.


Andrei, for the last part of our talk, I’d like to ask two broadly retrospective questions about your career and then one looking forward. So, the first thing I’d like to ask is, I’m an American. I interview mostly American physicists, mostly American counterparts of yours. I’m curious. In your intellectual development growing up in Russia, in the way that you came about studying the things that you wanted to study, do you see yourself in a particular intellectual heritage or lineage, and if so, do you see that lineage as specifically Russian?


[Chuckles] For scientists in Russia, we knew that there are some upper limits of our salary which we can reach. The most important reward for your work was not money but scientific achievements and respect of our colleagues. Of course, if you can become an academician, there is a fight for that because people then get some additional privileges, et cetera; their salary grows up. But basically, mostly it was about science. It was about finding what is true. Mutual respect to each other was based on it.

This is not always the same everywhere in the world because here the possibilities of success in science and beyond are unlimited. This is one of the reasons why I like Stanford so much, because here we found ourselves in the atmosphere that was very similar in spirit. When they invited me to visit Stanford 30 years ago, I have seen Lenny Susskind sitting in his old jeans on the floor and talking to students, and I liked it. Then he took me around the campus in his car. When he opened the glove compartment, it just fell down to the floor, because the car was old and barely running, but it was Lenny Susskind, you know? Who cares about the car? That was the same spirit, okay?


Andrei, we certainly can’t talk about all of the awards and recognitions that you’ve gotten over the course of your career, but I wonder. Given the field that you work in and given how difficult it is to convey the concepts, even to your own peers within the physics community, I wonder if there’s any award or recognition or membership that stands out in your mind as a symbol of ratification or verification by your colleagues and your peers that your work is really historically significant and valuable.


Well, for me there are two or three awards. The largest monetarily was the Fundamental Physics Prize, of course. It was huge and totally unexpected. But more importantly, this prize was given to people whom I deeply respect for their intellectual abilities. At first, many of them were string theorists, but because of the intellectual power of the people involved, the group of the laureates rapidly evolves and becomes much more diverse. It will not happen overnight because we are influencing each other, and our knowledge is not absolute, but we are awarding more and more people from astrophysics, from solid state physics, from every other branch of physics. I am very impressed by it.

But I should also mention the Peter Gruber Cosmology prize, which we received together with Alan Guth long ago, and the Kavli Prize for astrophysics, which Alan Guth, Alexei Starobinsky and I received relatively recently. When cosmologists give you a prize for cosmology, you know that real experts decide. Of course, as I learned being on different committees, it does not always mean that everything is objective and everybody understands what the prize is for, but that is how democracy works, and it’s amazing to see how sometimes a committee, consisting of people not always knowledgeable about everything (including me), comes to decisions, which, looking back in time, were quite reasonable. So yeah, we’re learning democracy the hard way.


Andrei, my last question looking forward. Because inflation… You’ve been involved with it for over 40 years at this point. There have been no competing theories that have knocked inflation off its pedestal. What are you optimistic about in the future as the theory of inflation continues to strengthen and as perhaps more observational data makes your original inclination seem stronger and stronger?


Well, it’s not exactly like we absolutely don’t have competitors. There are some alternative attempts to solve major cosmological problems. The most famous one is the ekpyrotic/cyclic scenario by Steinhardt, Turok and collaborators. Their first paper looked really interesting, so we decided to check it. And we found that instead of solving the homogeneity problem, this scenario required the universe to be exactly homogeneous from the very beginning. They claimed that they invented a new mechanism of generation of perturbations similar to inflationary perturbations. But the subsequent investigation by the best experts in the theory of cosmological perturbations demonstrated that their mechanism did not work. But the most significant problem of the ekpyrotic scenario was that instead of the big bang predicted in that paper there was a big crunch. In other words, they thought that they are describing creation of the universe, but in fact they described its collapse. This completely invalidated the original scenario.

I could continue discussing further evolution of this scenario, or the pre-big-bang scenario, or string gas cosmology, but I better stop here. Cosmologists do not really care about our personal attitude or ambitions, they want to find the best theory, they study everything, and then they vote by their feet.

I think that at the moment, inflation is the simplest and the most elegant way to solve many cosmological problems. So, inflation, or something very similar to it, probably must be right. But it is certainly incomplete, it is just a part of the story, which should be incorporated into something else. For example, there was Newton’s law of gravity, and then the Einstein’s general theory of relativity incorporated and generalized the Newton’s theory. So, what I really hope for is that there will be some theory which will put everything that we are doing right now in a more powerful context, and this context may involve something which we do not know much about, something like what Lenny Susskind and many others right now study here at Stanford and in many other places, like quantum theory of information, like where the universe came from and where the laws of physics were written when there was no universe, all of these questions. These questions must be addressed. I often say that in the simplest inflationary models everything could originate from a Planck-size domain. And then I think, yeah, but can one write all laws of physics in a single domain of the Planck size, or maybe one needs a larger domain to incorporate the genetic code of the universe. Or maybe what we call “laws of physics” emerged “on the way,” differently in different parts of the multiverse.

One does not need to buy everything that everybody else is selling right now. We are searching for something. I hope that in the end of this search, inflation will be a part of a more general cosmological construction, in the same way as Newton’s law of gravity became a part of the general theory of relativity, which, in its turn, must be extended to become compatible with quantum theory, perhaps by incorporating it into string theory. If this works, if inflation will be part of what survives, then it will be great. I’m not sure that I am going to witness the time of a final decision of what is great and what is not so great. Our life is short, but the universe is still young, so it has lots of time to study its own origin and structure. We are trying to do our best with our limited abilities, we are happy with what we already have, without being satisfied with where we are.


Andrei, it’s been a great pleasure spending this time with you. Thank you so much for sharing your insights over the course from 1987 and even a few years before. For the historical record, I’m so glad that we were able to update and get everything in the record for your work in the last 40 years, so I really appreciate this.


Okay. Thank you very much for the very thoughtful and well-structured interview.