My guest today is Prof. Thomas Hartung, MD, PhD. He is the former Head of the European Commission’s Center for the Validation of Alternative Methods (ECVAM), Ispra, Italy, and has authored more than 560 scientific publications and received 16 different awards ranging from the German Ministry of Health to Hellenistic Society of Toxicology.
Prof. Hartung is presently the Doerenkamp-Zbinden-Chair for Evidence-based Toxicology in the Department of Environmental Health and Engineering at Johns Hopkins Bloomberg School of Public Health, Baltimore, with a joint appointment at the Whiting School of Engineering. He also holds a joint appointment for Molecular Microbiology and Immunology at the Bloomberg School. In addition, he holds a joint appointment as Professor for Pharmacology and Toxicology at University of Konstanz, Germany; is Director of Centers for Alternatives to Animal Testing (CAAT, http://caat.jhsph.edu) of both universities. He is adjunct professor at Georgetown University, Washington D.C. and he is Chief Editor of Frontiers in Artificial Intelligence.
These past few weeks, image generators like DALL·E 2 and language models like ChatGPT are grabbing headlines. But Prof. Hartung has been working on what sounds like an equally, if not more, revolutionary research, namely computers powered by human brain cells, part of a new field called “organoid intelligence.” Prof. Hartung explains the concept and the research.
“Brains have an amazing capacity to store information, estimated at 2,500 (terabytes). We’re reaching the physical limits of silicon computers because we cannot pack more transistors into a tiny chip.”
In essence, Prof. Hartung is interested in two different things. One is in advanced cell culture, focusing mainly on the brain referred to as brain organoids, which is a really a breakthrough technology in recent years. And the second one is artificial intelligence, AI. And merging these two, brings them together to AI Augmented Intelligence. So the idea is, how far can we get by letting brain organoids for us do also some computational work.
.In response to my questions as to what's the most important thing he has learned about, about life, Prof. Hartung said, “It all happens between people. It is not about the next scientific paper, or the speech you give. It is about talking to others, inspiring each other and motivating each other to work for a goal. It's not something which is done when I'm writing my next article on the computer.”
If you liked this podcast
https://www.psychologytoday.com/intl/contributors/thomas-r-verny-md
My guest today is Prof. Thomas Hartung, MD, PhD. He is the former Head of the European Commission’s Center for the Validation of Alternative Methods (ECVAM), Ispra, Italy, and has authored more than 560 scientific publications and received 16 different awards ranging from the German Ministry of Health to Hellenistic Society of Toxicology.
Prof. Hartung is presently the Doerenkamp-Zbinden-Chair for Evidence-based Toxicology in the Department of Environmental Health and Engineering at Johns Hopkins Bloomberg School of Public Health, Baltimore, with a joint appointment at the Whiting School of Engineering. He also holds a joint appointment for Molecular Microbiology and Immunology at the Bloomberg School. In addition, he holds a joint appointment as Professor for Pharmacology and Toxicology at University of Konstanz, Germany; is Director of Centers for Alternatives to Animal Testing (CAAT, http://caat.jhsph.edu) of both universities. He is adjunct professor at Georgetown University, Washington D.C. and he is Chief Editor of Frontiers in Artificial Intelligence.
These past few weeks, image generators like DALL·E 2 and language models like ChatGPT are grabbing headlines. But Prof. Hartung has been working on what sounds like an equally, if not more, revolutionary research, namely computers powered by human brain cells, part of a new field called “organoid intelligence.” Prof. Hartung explains the concept and the research.
“Brains have an amazing capacity to store information, estimated at 2,500 (terabytes). We’re reaching the physical limits of silicon computers because we cannot pack more transistors into a tiny chip.”
In essence, Prof. Hartung is interested in two different things. One is in advanced cell culture, focusing mainly on the brain referred to as brain organoids, which is a really a breakthrough technology in recent years. And the second one is artificial intelligence, AI. And merging these two, brings them together to AI Augmented Intelligence. So the idea is, how far can we get by letting brain organoids for us do also some computational work.
.In response to my questions as to what's the most important thing he has learned about, about life, Prof. Hartung said, “It all happens between people. It is not about the next scientific paper, or the speech you give. It is about talking to others, inspiring each other and motivating each other to work for a goal. It's not something which is done when I'm writing my next article on the computer.”
If you liked this podcast
https://www.psychologytoday.com/intl/contributors/thomas-r-verny-md
SUMMARY KEYWORDS
brain, human, computer, people, system, supercomputer, toxicology, moment, work, AI, stem cells, life, research, science, intelligence,
SPEAKERS
Speaker 2 (74%), Speaker 1 (24%)
1
Speaker 1
0:00
Good morning. This is pushing boundaries, a podcast about pioneering research, breakthrough discoveries and unconventional ideas. I'm your host, Dr. Thomas R. Verny. My guest today is Professor Thomas Hartung MD PhD. He's a former head of the European Commission center for the validation of alternative methods and has authored more than 560. Probably by this time, probably it's 561 scientific
0:37
40. In the meantime, thanks for having me.
1
Speaker 1
0:41
I don't know how you ever find time to write all these papers, and he has received 16 different awards ranging from the German Ministry of Health to Hellenistic Society of Toxicology. Prof. Hartung is presently the Doerenkamp-Zbinden-Chair for Evidence-based Toxicology in the Department of Environmental Health and Engineering at Johns Hopkins Bloomberg School of Public Health, Baltimore, with a joint appointment at the Whiting School of Engineering. He also holds a joint appointment for Molecular Microbiology and Immunology at the Bloomberg School. In addition, he holds a joint appointment as Professor for Pharmacology and Toxicology at University of Konstanz, Germany; is Director of Centers for Alternatives to Animal Testing (CAAT, http://caat.jhsph.edu) of both universities. He is adjunct professor at Georgetown University, Washington D.C. and he is Chief Editor of Frontiers in Artificial Intelligence.
Welcome, Tomas.
These past few weeks, image generators like DALL·E 2 and language models like ChatGPT are grabbing headlines. But Prof. Hartung has been working on what sounds like an equally, if not more, revolutionary research, namely computers powered by human brain cells, part of a new field called “organoid intelligence.”
1:59
1
Speaker 1
2:03
So would you please tell me more about that?
2
Speaker 2
2:36
I mean, from all the different affiliations, I have. In essence, I'm interested in two different things. One is in advanced cell culture. And here we are focusing mainly on the brain. We call them brain organoids, which is a really a breakthrough technology in recent years. And the second one is artificial intelligence, AI. And merging these two, bring them together gives AI augmented intelligence. So the idea is, how far can we get by letting brain organoids for us do also some computational work. And this sounds like science fiction, and probably is in large part, because we are at the very beginning of this, but we see that these organoids can actually start to play the computer game Pong, or do similar things. And even though the computer, which is big brain equivalent, is far away, and perhaps we will never want it to be there. There's a lot of gain on the way which we are at the moment systematically attacking.
1
Speaker 1
3:50
So you have been, you have said, I think I have a quote here, which says, brains have an amazing capacity to store information, estimated at 2500 terabytes, we are reaching the physical limits of silicon computers, because we cannot pack more transistors into a tiny chip. And that is why you think that we have to go the way of the organoids Correct.
2
Speaker 2
4:18
Yeah, let's say this was one of our advertisement lines. I mean, the you have to compare here a reality, which is at the moment we we wouldn't be very happy if we could do what handheld calculator can do. And we are comparing really, two extremes. One is the highly optimized brain after millions of years of evolution and supercomputers. They're still an edge for the for the brain, but to promise that we will at any time have an artificial system which is outperforming a supercomputer is certainly a very bold claim. But the comparison still shows that biology has done it differently. And because it does differently also offers a potential. And whether we will really build them as a cellular computer, or whether we simply will learn how the brain does the trick, and then what we call neuromorphic computation, that we are trying to build our systems following this role model, these are two completely different ways to go.
1
Speaker 1
5:34
So I am somewhat familiar with Michael Levin's work at Tufts University. Also an organoids tiny little artificial animals that he has kind of produced. Is that similar to your work?
2
Speaker 2
5:57
Yeah, there's on the one hand, yes, this, let's say, it really was these two pillars of AI and the organ network, which were brought together here for the first time, which might make make the difference. And, you know, to, to grow cells of humans is something we can do since 1956, when the first human set to muscle and rockets. If you are working on brains, like I do, it's very difficult to pick pick people's brain in a non literal way, you will not be willing to, to make any donations for for my work. So we were really restricted in only being able to use animals or animal brain cells, until stem cells of humans were introduced. And they are ethically non problematic only since 2006. When Yamanaka found a way that you can take just a bit of skin or later blood cells, which are reprogrammed to become stem cells. So not problematic. And informed consent, the donor knows that you'd want to do research with it and what you want to do, typically, and then you can start forming organs. And it took until 2013, when the first brain organoids of human cells were generated. This was a big show, Nature paper, for the Lancaster. And paper from the canal bridge group in Vienna. We were at the time already two years working ourselves on brain organoids and had started publishing, but they had the big shot him. And it was a bit more than two years later, that we were able as the fourth group, publishing on brain organoids that we've able to mass produce them. So instead of each of these brain, organics being very large and looking very different, and most of them somehow bit rotten in the core, because oxygen and nutrients did not really enter sufficiently be produced 1000s Millions of them. And this was a big hit, because that's was the starting point for using them in toxicology. And in pharmacology, because if you want to develop a drug, you cannot easily work with things which look very different every day and every average each and one of them. And so we have been introducing really a new quality here and in since 2016, we have been publishing on one disease after the other with these brain organoids are helping us to work with human relevant models in a non problematic and fast and economic way.
1
Speaker 1
8:42
So is is there? Is there any danger of these organoids becoming conscious of themselves like becoming like little animals themselves?
2
Speaker 2
8:56
This was the internet instant reaction of people. I presented this at the AAA s which is the very big science journalist conference by the American Association for Advancement of Science in 2016. This was the first time we've been public Institute somebody's are they conscious and my line and it has been my standard line for many years following this was they are thinking because they are logically active, but they have nothing to think about. They have no input and no output and organic intelligence is exactly changing this. It is saying what happens if we give them input and output. So at the moment, I would say there is no risk of any type of consciousness in the foreseeable future. But in principle, we are enabling exactly the hurdle via making them possibly aware of their environment. And we know We'll have to start to define what is really sentience and what is offering what is intelligence? Yes, what is consciousness in the end? What is life? What is life life I mean that they live they fulfill all definitions of a living being they respond to the outside there they are metabolizing so, there they really show that they are living system but are they are perceiving what do they know about the environment? What are they capable of? It is really we really have to define consciousness, sentience and all of these things from its minimum, how much does it take to make a claim this is an intelligent system this is intelligence in a dish
10:52
can they reproduce
2
Speaker 2
10:55
um, brain cells are dividing only a short period in Organogenesis and then they are terminally differentiated. So, they would not typically grow again, though there is a small stem cell niche as we call it a few stem cells, which can be reactivated and sometimes repair things. But in general, it is an organ which, which is just maintained. And as we all know, coming to old age is then continuously declining and failing us, it is not growing new cells and allowing us to rejuvenate the brain.
1
Speaker 1
11:36
So, how does how does a computer really stack up against the human brain?
2
Speaker 2
11:43
I mean, in our paper, we make a comparison to the to the supercomputer, the number one supercomputer in the world, which is the frontier supercomputer in Kentucky, which, in June of last year, for the first time exceeded one axon flop, which is a measure of computational power, the number of operations which can be run, and one exaflop is exactly what a single human brain is estimated to deliver. We already quoted the 2.5, petabyte or 2500 terabyte, which is the assumed memory capacity of the human brain. Still pretty good. That's a mainframe computer in in a university. So this this rank order of, of capacity, you can you can also say the different way, the brain is about 1.5 kilogram so like, like a laptop, we know that our computers get twice as strong every second year, for the last 60 years Moore's law, you have to wait another 33 years until your laptop performs as good as the human brain. It shows you the distance we have from with respect to our brain.
1
Speaker 1
13:05
So I have I've heard a lot about the Turing test, which is to tell computers and humans apart, can you can you tell us a little bit about that? What is the Turing test?
2
Speaker 2
13:17
I mean, in the very early days of computers, and robots, Turing said, in essence, if you cannot distinguish a computer, and its answers from from a human being, then it is intelligent and tests passed the so called Turing test now actually, we identify a computer very quickly. If you use these new things, chat GPT so you sometimes might, might think, Oh, this is an intelligent human being talking to you. But it's very simple trick, ask him what the square root of two is, and this will instantly give you 100 decimals after the comma. And, and you know, that this was not done by a human, right? So we identify them with the outperformance it is, it we are but we're actually relatively close already now, to something which is indistinguishable from the human also, with respect to the, let's say, the outperform the the high quality information. So if you take GPT four, which just came out on the 14th of March, it is performing at the 90th percentile. So better than 90% of the students in a simulated bar exam for lawyers. So whatever you think about lawyers, that's a pretty good intellectual task. Yes. And it shows that these systems are really catching up. Mm
1
Speaker 1
14:50
hmm. So you started off in environmental health and toxicology right? Your career your academic career.
2
Speaker 2
15:01
You know, I'm there's two types of scientists, there's one type, which has an area of work and to get deeper and deeper and at the end of the Pope of this field. Yes, I've always been brought. I did study biochemistry and human medicine and mathematics informatics, I have collected as you counted five professorships in, in the introduction, which are all very different from the so I really cannot say this is my area. This is why I'm always looking for connecting areas, like in case of organic intelligence, AI and brain organic network, which is a such a new combination. While there have been people over the last 20 years trying to do brain machine interfaces, or have been using AI to analyze the EEG of patients, but using the two together is really the the novel aspect of it.
1
Speaker 1
15:59
So was there anything in your own background that you think contributed to your sort of more 180 degree interest in science?
2
Speaker 2
16:13
No, I've always loved all sciences. And like a spot. I was, I was already in high school or before middle school, I was doing these spheres and very different topics from biology and chemistry. And I'm simply my type I'm, I'm fascinated by many things, and, but I always like to convert it from trying to understand to trying to do something with this knowledge. I'm trying to build something from what I have learned. And man, I see such an opportunity to do something which could have an impact, I love to jump on it.
1
Speaker 1
16:53
So what do you think will be the impact? Or what do you hope will be the impact of this work in the long run?
2
Speaker 2
17:00
I mean, the important message is that bio engineering with stem cells has given us now brain systems, human brain systems, which we can use for directly for studying aspects of neuroscience. If you have a system, which is showing some learning some memory, you can just study learning and memory and human system without using monkeys, or the difficulty of testing something on humans, the non mirroring of human physiology by rodents. All of this aside, you have this very simple system where you can test that you can manipulate all the stuff we scientists love. Yep. And we can find out whether our assumptions and and our textbook knowledge about learning and memory is correct. And drive it further. The second element then comes to where I'm mostly working, which is, can we use such a system to identify chemicals which are producing autism? Can we use this to identify what is possibly contributing to Alzheimer's disease? Because it is not just the aging of our society, there's something more about these diseases are dramatically increase it. And then you can do the next step you can ask can be prevented can be cured? Can we develop drugs to help with this and and these are all near term goals. These are goals where we essentially we are working on this already. We are producing brain organoids from Alzheimer patients from autistic children, and we are trying to see how do they behave in comparison to those from healthy donors. And the third aim is then the computing site. And this is certainly the most visionary of all. It is something where we have a long term perspective necessarily. And where, as I said, we might just learn how to build our computers better from it. But it could very well be that at some point, we decide a large organoids perhaps several kilograms large could add something to our silicon based computers as a component to which is helping and supporting it. Yeah.
1
Speaker 1
19:24
It's very, very interesting. Very interesting. So are you are you are you presently working on a book?
2
Speaker 2
19:36
No, I don't. My former mentor always said, write books after your retirement. Okay. In our fast paced scientific world, we are trying to get things out as quick as possible. You have to see AI over the last 1012 years has been doubling in capacity every three months. If I write a book, you know it, you have that. So it takes you a year after writing until it's on the market. So everything would be outdated. If it's fast paced field like, like mine. So I'm really trying to get things sometimes even out without going through these long peer review processes of science, just putting out as my personal opinion and saying, Hey, guys, I would like to share something with you, before it is outdated.
1
Speaker 1
20:28
Well, speaking of that, so did you. Did you meet any obstacles to your work? Did you run into any resistance to this kind of work at any of the universities that you are associated with?
2
Speaker 2
20:45
No, not at all. And I think this has to do first of all, was the fact that we from the very beginning, versus bioethicists, because ethical questions would, if at all, be the ones which would be raised. So we teamed up here in Hopkins was the Burma Institute for bioethics, which is really a leading organization in this field. They have been part of of many projects at this moment, they already serving a general population as a about his research, when people stopped saying, oh, perhaps you should think about this twice. And they asked, and, you know, this is done with informed consent. So the people who gave the cells agreed on this, and this is done to find cures for Alzheimer, we want to know, does this change perception of people? Yeah. We don't want at some point to come out with a modern system and say, here is the product and they say, right, understand this is very important that we have an ethical discussion. And we have the ethicist sitting with us, but we plan the experiments, they come to our labs and see what we are really talking about. And then this puts a lot of things into perspective, if you see these tiny brain organoids. In which often have about the number of of neurons as a flyer. So you don't believe that this will be a system? You can talk to you if you've
22:09
not yet Not yet. Are your parents still alive?
22:14
My mother is Yeah.
22:16
And what did your father do?
2
Speaker 2
22:19
Well, my father was an accountant and see a CFO of a company building.
1
Speaker 1
22:27
Did your parents understand your work? Did you talk to them? Were you able to talk to them about it?
2
Speaker 2
22:34
No, they don't, don't really don't think they understand. They understand some of the big lines, and they occasionally read an article. So my mother is proud to see me in the German newspapers or occasion. If it's anything in German, I send it to her. But I don't I don't think she would get much more than that. I'm trying to replace animal research that I'm respected scientist, which makes me very proud.
1
Speaker 1
23:01
Right and light and rightly so. Do you have any children? Yes, I do. How old? Are they?
2
Speaker 2
23:09
Um, we have a 16 year old and a six year old.
1
Speaker 1
23:13
Okay. And the 16 year old does he understand the see understand more your work?
2
Speaker 2
23:19
Yeah, but he does and but his interest is clearly more into art. Also, have more discussions about movies, and how they are made. Have some politics, but he occasionally gets something of our discussions, but not more. But his mother is actually part of my team. And she's the first author of the paper. Yeah, she's doing the brain organic research in my team. Oh, so
1
Speaker 1
23:47
you work together? Sometimes. We are absolutely everyday we,
2
Speaker 2
23:51
you know, my way of retaining talent is to marry. And so I changed her job profile to 24/7 I often say, and we are, it is wonderful to work together and discuss ideas. Very strange times of the day.
Did you ever disagree?
Oh, yes, absolutely.
24:15
And this is good thing. Scientists who are agreeing are boring.
24:22
That's true.
2
Speaker 2
24:24
different perspectives. And this is certainly very important part of it.
1
Speaker 1
24:29
Well, it sounds as if you were working very hard. How do you relax?
24:35
I mean, firs,t sleep is overrated.
yes.
2
Speaker 2
24:43
I love jazz music and photography. And if I really have a lot of time, which does not happen often I like some magic.
1
Speaker 1
24:54
Oh, really? You do magic magical tricks.
2
Speaker 2
24:58
Yeah, I'm mainly in mentalism. Yes, because it is. So just reading minds making predictions, all these impossible things and because I think it's just the opposite of science, yeah. Producing purpose and to entertain. Yeah. And it tells us a lot about how we mislead and are misled. I really think it helps to understand also the logic and the logic of the scientific machinery. Hmm.
1
Speaker 1
25:32
Interesting. Interesting. So in terms of your life, can you tell me how old are you approximately 50, 55?
25:42
Thanks a lot. I'm turning 60. This year
1
Speaker 1
25:46
60? Well, you look much younger. So in your 60 years,
2
Speaker 2
25:50
I was always perceived as much older. When I was young. Yes. seemed not to age. Well. I was considered early in my late 20s. When I came to university.
1
Speaker 1
26:06
Well, I had the opposite experience, I was always perceived much younger, which is difficult for a psychiatrist, you know, like, who is going to? Who's going to trust the teenage psychiatrist, right.
2
Speaker 2
26:18
That's an interesting one. Yeah. So we should compare our genomes to find Yes,
1
Speaker 1
26:22
yes, we should, we should. Tell me about a peak experience in your life when you were really, really, really happy.
2
Speaker 2
26:34
peak experience. I think the happiest moment of my life was when my daughter was born. Yes. Just, it wasn't a moment of complete fulfillment and satisfaction. And you know, the best moment was not the moment when she was born. Yeah. The brother came and saw her for the first time. Yes, this moment, his fascination triggered all emotions in my life.
27:03
Did you cry?
2
Speaker 2
27:04
Yes, a bit. And I actually have a picture of this encounter on my Mac here, you see, oh, yes, I see. Yes. Oh, sorry. I've got this right. Now
27:14
I see. I see. Yes,
2
Speaker 2
27:16
this was the very moment because I share it so much. Cherish it so much. I'm, I kept it and have my daily tree from
1
Speaker 1
27:24
it. So you have one daughter and one son.
2
Speaker 2
27:27
Yeah, Vanya was already five and he came into my life is wonderful. And I love him as dearly as my daughter. I just didn't have this very experience with him. Right? By I can share this without being unfair to him. There's no difference in the affection I have for him. But the one moment here, when the four of us together, just came together as as, as a family with this little thing
1
Speaker 1
27:59
might become Right. right. In terms of your life, in terms of your 60 years, much of its spent in academia. What's the most important thing you have learned about, about life, about yourself about yourself?
2
Speaker 2
28:23
I think one, one thing is you get more humble, especially if you're working in the medical field. And if you see how much we don't know, and how much we can't cure. And so this was a very important experience for me after a lot of these scientific research going to work two and a half years in hospital. And the second thing is that it all happens between people. It is not about the next scientific paper, or the speech you give it is about talking to others, inspiring each other and motivating each other to work for a goal. So it is a highly social thing. Political thing in the sense of policy of, of society. It's not something which is done when I'm writing my next article on the computer.
1
Speaker 1
29:15
So that takes me to my last question, I always ask my guests this question. What does being human mean to you?
2
Speaker 2
29:30
Being Human for me means to be aware of this, the people around you and not falling into the trap of just seeing as as a technical problem I'm solving. Everything I'm doing is done with people who are highly complex and I need to understand them their motivation, but I can do for them too. Make it a win win situation for all of us. We will work together.
1
Speaker 1
30:05
Thank you. That's, that's wonderful. Thank you. Thank you so much. I think I will let you go because I know how busy you are. But I want to thank you for a most informative and human interaction here. I want to say that my guest in two weeks will be Dr. O'Leary. She does research and writing on prenatal parenting and pregnancy. Her particular specialty is the study of parenting after the loss of a baby. Its impact on both parents and siblings, and a child born after the loss of the baby, please tune in. And once again, Dr. Hartung, thank you so much, and good luck.
30:45
Thanks for having me.
30:47
Take care. Bye bye