Review of Scientific Paradigm Shifts

Heat. Heat. All right, let's let's begin then. So, welcome everyone to this event at the Science Circle. We are doing a 2021 revisited panel. And so 2020 was a year that was highly active for the science circle on a lot of very interesting topics. And as a way of, you know, reflecting and spending a little bit of time revisiting and updating from that that year, we do these revisited panels. And so today, uh, the two main talks that we were going to do a revisit on was a presentation by Jeppi Longo on art and physics, but uh, he is he has not arrived yet. The other one was, as you can look at Matt, sorry, look at Matt's screen behind us, and that's showing that we had a panel back in 2020, a uh a conversation about paradigm shifts and in the context of and his uh book, the structure of scientific revolutions. So the three speakers sorry right the three panelists in that case were Matt as moderator who is here again today to give us an update and then Jeppi Longo and actually myself. So if Jeppy uh does not show up then I will give a little bit of a conversation and some uh some thoughts on paradigm shifts uh not as moderator but more as as a panelist. And so, uh, what I will do now is turn the microphone over to Matt and he'll give a little bit of an introduction about himself and then give you some updates on thinking about paradigm shifts as well as a little bit of a review, right? A little bit of review. All right. Uh, thank you, Stephen. Can everybody hear me? This is, uh, Baragon Betts, also Matt Burr is my real name. Um, so hopefully you can hear me. Uh I did post in uh nearby chat just now a link to the uh to the OG panel discussion uh which uh was uh you know if I say so myself pretty darn interesting. Um uh you know uh you know for example uh you know we talked about topics like uh like diseases like chalera and malaria sort of general biology you know changes in uh like big jumps in progress uh in in those topics. Um uh we talked about science as a profession. We talked about falsification. Uh that you know kind of the uh uh how one uh I think it one of the u I guess tenets of science that has become adopted recently is the ability to be falsified. Whether a proposition can be falsified if it can't be falsified you know it's uh it can be sort of preserved. Um so we talked about of course Dharma's theory of evolution. We talked about dark matter. Uh we talked about um um uh what else? Well, let's see what else. And then finally, we had a long discussion at the end of that about how overwhelming science can be. So um uh and I think I think we can all sort of relate to that. Um but uh basically a uh just also just a quick primer here that uh a paradigm shift um is kind of a uh kind of a sort basically a permanent and sort of radical change in the way science uh or I guess not even science but just the way humanity um sort of understands the world. And I think the classic example of that is the capertic revolution where we uh universally came to understand that the earth goes around the sun um for example um and also uh you know another example might be Darwin's uh theory of evolution uh about how the how species originate right so those are those are paradigm shifts you know you move away from from sort of supernatural creationism to a science um based understanding of how species originate, right? So those are examples of paradigm shifts. So um and that's what we talked about like five years ago. Um but let's take a look. So things have changed quite a bit. Let me go to my next slide here. Let's just kind of get into it. Uh there we go. So I apologize apologize for all the text. Um, uh, but I'm hoping you can just kind of maybe zoom in on this slide here. Uh, I do think that this is a, uh, a pretty intriguing, uh, uh, list of, um, of topics that we can discuss. And by the way, feel free to, um, you know, put your comments in nearby chat or or even, you know, I don't even mind if you join in by voice if you want to. So, we can keep this, you know, fairly informal, I think. Um but uh I'll just quickly kind of run through the the topic uh the headline topics here and then we can dive in uh to more uh granular discussion of each of these. So um uh u so just quickly modern paradigm shifts represent fundamental changes how society science and industries operate moving from established norms to entirely new approaches. Um, okay. So, we're going to be looking at uh things like of course AI, you know, that's going to be inescapable. Uh, also um the uh the growing uh significance of uh electric transportation and just I would say sort of green uh energy sources in general. Um also you know the transformation of Hollywood to streaming uh platforms and digital content I think uh is still uh uh still making a huge impact uh in our daily lives. Now also remote work I think has really transformed the economy and society since co um and of course we now have tools to be able to to do that. Um and then here again sustainable and renewable energy know that's relates to the electric transportation I think but so we can get into that also we have uh oh this idea cloud computing and SAS service as a software as a service is SAS and um you know this kind of uh well uh this kind of we don't own anything anymore. we're just renting, you know, all of our all of our content. Uh, and that has some interesting, uh, I think uh, implications. Also um of course the emergence of uh social media, the attention economy um you know how how you know how do we um I guess where where do we position the importance of social media in you know especially sort of in competition with our actual real physical lives. Um and then um finally uh excuse me the last one here scientific understanding. So um I think uh the example given here on the slide is uh the fact that we discovered that there is a bacteria that causes ulcers. Um, but I think what I'd like to get into here, frankly, would be a discussion of the current Maha movement, um, of, you know, make America healthy again as as embodied by RFK Jr. in terms of his understanding of science. We are now kind of in a under a regime that has a radically different understanding of what science is than I think we have been comfortable with in the past. All right. So, um uh um let me So, uh let me go back and take a look at our I want to go back and look at some comments here in nearby chat. Uh we don't really own literature anymore. That's right. Shiloh points that out. But yeah, and then Sizzy, uh the EVs actually preceded gas engine cars. That's exactly right. Isn't that interesting? Yeah. Uh and uh um and a day points out that the UK is going to ban children under 16 for having special media accounts. I think uh didn't uh didn't Australia already do that actually. So um misunderstanding of scientist the 1830s. Yeah. Yeah. Well, we're going to get to all of this. Uh okay. Let's let's start at the Yes. Thank you, Saj. That's right. Australia did do it. But let's start at the at the top of this uh um of our list here and let's just get into AI. All right. Uh in fact, you can tell actually that this slide was prepared by AI, right? Uh so uh you know u so I I did a prompt on this and then you know basically uh made an image of the results, right? So that's uh I think I think fairly common that everybody is doing this now, right? Um and uh I you know I don't know if I have anything particularly uh insightful to talk say about artificial intelligence and generative AI um uh you know that uh any more than what anybody else has. You know I have my anxieties about it. Um I you know it is um uh you know it has been interesting uh for example this spring uh when um the tech people were giving their uh commencement addresses to new graduates at universities and when they mentioned AI they got booed. Um and I think this is because so many of these students spent four years you know really to prepare themselves to go into the tech industry only to discover that uh you know their their jobs are being rendered obsolete by artificial intelligence. Do you remember? I think maybe it was related to COVID, but there used to be kind of a a catchphrase of, you know, learn to code. Um, you know, just code, baby, code. Uh, you could always, you know, learning to code was kind of considered sort of a a um a a a golden ticket to being able to get a job. But that has just completely fallen into the trash bin because AI is now um can now, you know, write code instantaneously, right? And this is going to just completely challenge the uh the employment market for new for new grads. And of course, it's a it's a huge challenge for creative fields as well. You know, I have a particular Yeah. Matt Shack Max uh says that the the code doesn't always work and that I was you know as I as the words are leaving my lips I had that same thought which is that the code doesn't always work and you know I have this experience in my in my own profession uh you know I u I do patents and trademarks and uh a lot of my clients now are sending me you know AI generated content um and they are also using AI to help to quote unquote help me um negotiate with the uh with the patent office. Um but but what I often find is that you know the AI doesn't understand what a patent is. It doesn't really understand what a claim is. Doesn't really understand what an invention is. Um and so the the fact that so it's token predicting capacities don't quite meet the moment. So, I have to do a lot of um in some ways it's almost more work because I have to go in and kind of really carefully look at the content I get from my clients and edit it uh pretty closely and it's it's almost more work, you know, it's a little bit crazy. And then also uh I'm doing my uh CLE this month, you know, for continuing education. Uh, and I listened to um a presentation recently about copyright for AI and um uh that um that it is a huge challenge for the copyright office. Um just basically quickly that uh the you know the point is that only humans can create copyrightable works of authorship. Um so a machine cannot own a copyright. Um and uh also the um the copyright office will if you submit kind of a mixed work like a a maybe um a song or a work of um of visual art uh that you know was kind of at the product of a an AI prompt. Um, the copyright office will either reject that or it will actually parse out which elements of the work you've submitted are yours that are authored by you and the and you will have to disclaim the portions that were generated by the AI. So, it's very very challenging and I also feel that especially in this particular Yeah, exactly Shiloh. Uh it's also um uh uh I also fear how now I think this is the proper position for the copyright office to take uh right um but I do fear that especially this present administration is very very chummy with the AI community and I think and and also you know Hollywood I think there's going to be a huge amount of political pressure for the copyright office to grant copyright right to AI generated works. And I think this is a slippery slope to sort of chaos. So, we'll see what's going to happen with that. But I think it would be I I just don't know. I just don't know how we're going to make that work. You know, you can't uh you know, I mean, are we if you can say that a machine can own a copyright, are you can say that a machine can enter into a contract? Can a machine enter into a license agreement? I don't know. But my instinct says no. No, it cannot. All right. Well, so anyway, I can now that I've kind of gotten worked up about that, uh I we could probably go on for for a long time about it. Uh but we do have a lot to talk about. So I just wanted to basically just kind of uh um uh vent about my own personal work concerns about AI. And hopefully you found that interesting. But uh let's let's kind of touch on some of these others here too. Um so let's talk a little bit about uh electric transportation EVAs and stuff like that. Um again I don't have a lot of tremendous insight about this but I do think that it's extremely interesting that China um has made a huge commitment to uh electric uh vehicles. um you know they have entire some cities in China have entire fleets of electric buses for example uh they make very they make really lowcost high-end electric vehicles that are simply unavailable here in the west um you know I've seen you know you know maybe they're propaganda but I have seen Tik Toks that sort of show uh you know really amazing Chinese uh electric vehicles and things like that so Um and I just feel like again we're in a we're in a particular a particular moment where you know we are dropping the ball uh in terms of keeping up uh competitively worldwide uh in this arena of electric transportation um which is weird because even uh you know even Elon Musk is invested in electric transportation. So it's very I'm very very confused about uh uh Elon's uh sort of um relationship to this particular topic. I don't quite understand why he is uh it seems to me to be supporting policies that are detrimental to his own investments. So if anybody has any insight about that, please feel free to share them uh because because I'm kind of baffled by it. And then moving on, we have streaming and digital content uh which I think has completely transformed uh Hollywood and uh and television networks. Um it is uh like I think this idea all of this digital stuff you know has just recently in the last year just led to um a series of cascading um uh mergers of uh streaming platforms and traditional um television networks getting kind of bundled up into streaming platforms. Um this also again seems very chaotic to me. Um and um and it is troubling that we don't really own anything anymore. So um I u I again I don't really have any solutions. Uh hang on a second. I'm just going to go and check on the uh on the comments here in nearby chat. Make I don't want to make sure I don't neglect that. Um, >> we have says it's not really possible to keep up >> with the largest and so to be the most powerful country in the world, the EVU market is huge and highly competitive. That is absolutely right. So, um, and Nurero asked you guys, you guys really got into the AI thing and Neuro asked uh um, who owns the products of AI? That's right. So, um, and I don't know if I have a clear I don't know if anyone has a clear question to that. I do think that, um, typically, um, the person who caused the AI content to be created probably owns it. Um, I think it is improbable to me that it will be found that the machine or the algorithm itself owns the product of AI. But I think I think in this I almost think that ownership in this context is a little bit we may need some new concepts because I mean in my mind things that are generated by machines are just in the public domain like they're not owned by anybody. Um, but maybe I'm naive about that. And I think and it is I think the economics of having like it's obvious that all future content can't be in the public domain, right? I mean, I do think there is a proper role for the ownership of authored content. If you write a book, you know, you should own the copyright. If you write a song, you should own the copyright to that song. So and so forth, right? Um so um so I I mean I don't know what the solution is to these things but but these are we are clearly in the midst of a paradigm shift uh with all of these uh questions that are being raised right so >> so there's also an interesting question from Shiloh >> so Shiloh also mentions certain diseases generated vaccines well you know right that actually is a good point I did want to make sure I touched on that that there are some some fantastic um uh you know legitimate uses of AI you know the one that I think is the one of the most interesting is the uh what's the one that um that predicts uh protein folding uh so that um you know it will generate uh all of these various candidates for the way uh a string of amino acids will fold um and that uh this you know is extremely useful for us for devising uh new kinds of peptides that are reactive. We can we can sort of customize um you know active sites within a protein protein. We can build proteins with custom active sites and things like that. Um and the ability of uh AI to uh you know to sort of sift through you know just tons and tons of data instantaneously to generate this output uh I think is uh is fantastic uh you know and it's also used in astronomy you know for um uh for analyzing just the tons and tons of data that comes out from all of our new telescopes and things like that. That is a fantastic use of AI I think to just help us manage vast amounts of data. But I will say that you know uh that the value of that uh AI product is in you know what it allows a humans to do with it. How it allows us to analyze that output. So it doesn't have I don't think that the AI content has value in itself other than uh how it helps us you know make decisions about what we want to do or what we want to look at or what we want to make. So um all right I'm going to move on here because we're almost at the half hour point. I don't want to I didn't want to consume the whole hour here. Uh so um uh so let's just quickly touch also on remote work and digital collaboration. Uh you know uh I work almost exclusively remotely now and I have clients from all over the country all over the world really. So um uh so I think remote work has been a huge transformation. You know the downtowns of cities are becoming um empty. Um the the the buildings downtown in downtowns are being repurposed. Um it's it's just uh incredible really. Um and uh um uh I don't know. I mean most you know uh like you know I live in Austin and uh we have just a tremendous amount of of construction building construction going on but it's almost all residential. there's hardly any office uh office space being built anymore. So, I think that's interesting, too. Yeah. And Shiloh mentions it's a it's a change in tax space as well. Yeah. I do wonder like whether like how will um property taxes be affected by the um by the abandonment of uh downtowns and cities, for example. I don't know. These are all these are all these are just huge huge changes in society. And let's touch touch base quickly on sustainable and renewable energy. I don't know if you guys caught it recently. There was a uh a Senate hearing by uh one of the administrations um I forget. I think he might have been uh in the energy department. Josh, I wish I had remembered that, but um he was being challenged about wind power and and he was going on about how pointless wind power is. And I think that uh a lot of these um uh people now think that that the wind itself is the energy that for example like we're using wind to like push cars down the highway or something like that. You know, I don't know if they fully understand that the wind is used to generate electricity. Do you think the wind power blows? But uh but I you know, I I don't know. I think they I don't think they understand that we used energy that we used wind to convert it to usable energy. It's it's a it's very very strange. They also don't seem to understand that that batteries exist. So they seem to believe that a lot of these green energies sources um are only available when that source is happening that we don't seem to have the ability to store energy. So this is all very confusing to me. I don't quite understand that. But at any rate, setting that aside, I I do think that uh um uh that I think sustainable and renewable energy is still um I don't know if it really qualifies as a paradigm shift. Now, I think it's on the cusp of being a a paradigm shift. I think there is the potential of a paradigm shift in the way uh we we uh provide energy. I don't know if we're quite there yet. Uh and yes points out such batteries already exist in China. That's right. Uh, you know what I remember many years a not many years ago but I don't know five or 10 years ago reading a um an interview in the New Yorker with the CEO of Exxon and uh in that interview and I don't remember who the man's name was but he said that Exxon considered their biggest competitor was batteries. Um, and this was, you know, maybe five or 10 years ago. And they recognized that again that this uh um sort of paradigm shift of uh of um of renewable green energy uh was inevitable. This is the message I got is that they recognize that this is inevitable that we are going to have to have uh renewable energy and have to have ways to store it. So, so maybe I I mean I don't know if it's paradigm shift yet, but I do think it is inevitable that that paradigm shift is going to happen. Okay, let's uh let's again let's just move along here to cloud computing and software as a service. Uh so uh again this has to do with uh not owning anything. you know, it's um Spotify and iTunes and Netflix and YouTube and all of that stuff. Um, you know, uh, you know, even when you I don't know. I mean, I feel like I have uh purchased the um Oh, actually uh Stephen, are you trying to talk to me? I just noticed. Uh, >> well, there were a couple times where I pitched Are you hearing me now? >> Yes, I do hear you. Thank you. >> Yeah, there were just a couple times I was trying to direct the conversation and and you just kept talking. So, uh, >> I I did not hear you. I do hear you now, though. So, thank you. >> Okay. >> Uh, yeah. Uh so I think what I wanted to well let's before we continue on with cloud computing uh there were a couple questions in the chat related to ownership of products that I think is kind of a we should pause for a second to look at the distinctions that my understanding of trademarks and the way the patent office views AI generated content is that because they can't be sure of tracing the ownership of contributions and they can't be sure that there wasn't a contribution to an intellectual property that really should be based on what AI found but then didn't give credit to that the patent office right now is not granting um uh copyright to certain AI generated content >> and so that's one aspect of it but I think but the other question that came up from Shiloh was this idea that oh if a company is using AI to generate some sort of patent patentable type of invention. That that's distinctly different in that the cop the patent filing should describe and have um a reproducible way of talking about some invention. And the as long as the invention itself has novelty that is you know there's no prior art to what the thing does and that has novelty and value then that should be granted regardless of how one generated some sort of background understanding or information related to it. So I think that that's something worth addressing that there are different purposes of those two of filings. >> Uh I agree with you. I do think that uh an AI assisted or AI generated invention for uh for uh for a patent uh is that should not an issue because authorship is not really an issue with whether something is patentable. Um that it does that relates to sort of the ownership uh question. I mean, I suppose uh some parties might uh um might raise a fuss about who who the inventor is or for example, who can be like who has the legal rights to assign the ownership of that patent to a corporation or to license that patent to uh other users, things like that. um uh you don't so so you know the so the owner of that patent I think and also so and I mean here's kind of the the the dilemma is that you know so the AI generates uh let's say um drawings uh and a written description of some apparatus and then so but how does the AI transfer the ownership to you so that you can file a patent on it. Um, so that is a little bit uh uh unclear to me. In a way, you might you could almost say that you discovered the invention, but of course I don't know. Um, but that but then does the does the invention become sort of a uh just an algorithm like a mathematical formula that just you know that that the patent office considers unpatentable? So that's kind of a but so that's kind of a whole minefield about that uh that I'm not sure how that's going to get sorted out. But I do think there will be a tremendous amount of pressure, political pressure for at least corporations or business entities to be able to own inventions um primarily sort of devised by AI. Um, so however that gets sorted out, the law will somehow be able to accommodate that, even if it has to kind of twist the the logic of the law a little bit to accomplish that. Um, and I do want to make it clear that uh, patents and trademarks and copyrights are all distinct forms of intellectual property. Um, so a patent is just something that is new and useful and not obvious. That's all. Right? As long as it's if it meets those criteria, you should be able to get a patent on it. And not obvious has a specific meaning in the patent law, it basically means that the examiner cannot come up with any combination of prior art that uh will satisfy all the elements that you have claimed in your patent applications. That's what obviousness means. It may not have been obvious to you to combine all of those references, but the mere fact that those references exist and the examiner found them makes them obvious. A trademark is for promoting your goods or services. A trademark completely different from a patent even though they're in the same office of the government. They're totally different. Uh a trademark has to do with your trade, with your business, how you know the the brand name of your product, things like that. And uh you want to protect that you know that no one can pretend to be you by coming up for example with a you know like remember in coming to America where the guys had the the McDonald's ripoff the McDougall's restaurant might have the golden arcs not the golden arches so it's totally different but anyway um so trademarks are very uh different and they they mainly have to not be descriptive and they can't conflict with another another trademark. You know, in other words, in terms of being confusing about who owns them. >> Baragan, can I ask you something then about trademark and patents if Okay. I I used a pharmaceutical company. All right. And so there are solving issues with like I just posted um AI generated um you know uh solutions for uh antibiotic resistance or certain infectious diseases. Does the company apply for the patent first and then it goes to FDA approval or does it go through FDA approval and then they apply for the patent? >> Uh you always apply for the patent first. The the patent application is the first thing you do mainly because an element of patentability is novelty. Your invention has to be new. Um and um so the problem with waiting too long to file for your patent is that it will stop being new. um you know for example uh you don't want to wait you know for example if you apply to the FBA FDA for approval you then you have to disclose that to the FDA you have to disclose your invention to the FDA that might make it public now it's not new anymore you've ruined your IP rights to it right so you always want to file your patents as soon as you possibly can even if it's only as a provisional patent that's the that's the general sort of the my the conservative answer to that question. Um, finally, I do also want to point out that copyright is also very different from patent or trademark. Copyright protects works of authorship. So, you are the a you're not the inventor of the work. You're the author of the work. If you create a sculpture, if you write a song, you make a movie, you're the author of the work. And that's what a copyright protects are works of authorship. Okay? And um when you own the copyright, you have the exclusive rights to create derivative works from your work. If you write a novel u then you own the derivative works for example that's how you are able to license your novel to Hollywood to make a movie out of it. The movie is a derivative work of your novel right and as a copyright owner you have the exclusive right to make derivative works or to license others to make derivative works. So, so copyrights are are very powerful. Um, and uh and actually trademarks are very powerful too. You know, I think of all the forms of of intellectual property, frankly, I think patents are the most problematic to be honest. So, anyway, I didn't mean to get into a whole thing about that, but but I do think that, you know, all of these sort of IP issues are pertinent to discussing paradigm shifts in a sense, right? I mean because in a weird way uh these um uh uh the law has already kind of thought about a lot of these issues you know u at least from the from the pre-paradigm status of them right so when you when you are in the midst of a paradigm shift one of the things that is affected by the paradigm shift is the law how the law is going to react to it so that's all quite interesting I Okay. So, um anyway, um so we've got about 15 minutes left. Stephen, what do you want to uh how do you want to sort of proceed to do I can finish out my topics here, but um uh was do you want to see if Jeppi is going to have a should I leave time for Jeppe to uh have uh some you know presentation also? How do you want to proceed? uh Jeppi has not uh come online or made contact with me and someone else is checking with him on Facebook. Let's um let's uh switch over for just a second to a little bit of some commentary I want to make about paradigm shifts and the scientific process. >> Uh good. And then I think what we'll do is open up the floor to some questions and see if anyone has and just some thoughts because again let me say let me say two things is that for when we think about paradigm shifts you know the the vernacular of the term now applies to changes that are social changes that are economical uh changes that are cultural as well as technology and science. And so if people would like to reflect on on what that kind of means for society that just like what described, a lot of times paradigm shifts come up against a lot of resistance. And when we see that in the world, where do we see resistance? Do we feel resistance to it? But there actually lots of times in culture, unlike science, where the this paradigm shift can be very rapidly embraced and people can get very excited about it. Although in some cases it turns out to not be that much of a shift. So I think um as as we start to close out this conversation I would like for the audience to think about some of their thoughts or feelings or experiences with what we call paradigm shifts and make a contribution. I do want to come back to one specific thing which is this and you have them listed up here as scientific understanding that there are ways I think Jeppy was going to talk about this because he did at the original panel that when we understand science the way the process has worked is that we usually have an idea we construct some experiments and we generate large overarching theories that are really just models of the world and then science uh as we do science they either fit that model that we have or they don't and then if they don't that's why we try and change the model and what's interesting is that with AI or some of these other large computational ways of doing it we don't have we can actually use those tools to generate a variety of options of a theory and to take uh the the best use some degree of of fact fitting to say well what actually is the most likely theory behind something behind all these options. Now, he was more of a a theoretical and cosmological physics type of um scientist. So, that's I think in a space where we're still trying to figure out what the nature of the universe is. there are still very large competing hypotheses about the age, the origin, the science, the components of the universe that there's still a lot of interesting space in there where computational power can give us new or different insights. Uh but I do want to come back to um in the practice of science, how does AI really it does change the science profession in certain ways in that you can take very large data sets and try and sort through them to find things that correlate to what you want. So right now I am a I'm a molecular biologist at a seed company and some of the there are a lot of startups but I think large companies doing this too where they will use AI and they'll just generate lots and lots of genomic data expression data protein data metabolomics data and then do comparisons across species to say hey can we find a gene that can that is likely to convey some sort of a useful trait if we were to turn it into a biotechnology product. And so that amount of shifting and sorting through data that you otherwise couldn't do individually is a really interesting place where um you can do some early discovery to get leads. And I think the comparison here is most like to a a journalist is that you get leads, but you still have to do a lot of pounding the concrete to and you know doing lots of pipe heading to actually do experiments. And a lot of times those experiments are still something where the art and the the inside of an individual human being actually makes the biggest difference. Now there are places where you can actually automate an entire workflow where synthetic biology if you're looking to change the properties through mutagenesis and randomly changing stuff and sorting through there are places where some of that science becomes automated but the it hasn't changed the fundamental idea that someone comes up with a hypothesis wants to do something different and test it. So that's still a very you know human agencydriven driven process and so I think that's that's interesting as well but AI is making a big impact in terms of how we do and conduct some of the science. Yeah, absolutely. For like the the protein folding example I gave where we have where you know the AI will generate tons of of of plausible AI folding examples that we can then sort through to see which ones really work. >> Yeah. Yeah. That's alpha fold. Uh that's the Nobel Prize >> and David Baker was also involved in um aspects of using AI for protein structure prediction. And so, you know, that's an amazing thing like you can we now have this ability at a much now again there's a trade-off, right? We have a fairly reasonable guess that the structure of a protein that we're trying to work with that alpha fold is doing a fairly good job of predicting it. Uh, but you still have some uncertainty that in fact it wasn't it wasn't experimentally demonstrated to be the structure you want. So you get a lot more ease of of likelihood of things that you do get to work with, but there can always be this bit of uncertainty. But it is amazing. I' I've used it. It's pretty amazing. >> So yeah, >> that's cool. >> We are hitting 10 minutes left. So >> let's survey the audience. >> That's a great idea. And I did also just quickly want to point out that I do think AI has been amazingly quickly adopted by the public. Uh people love to talk to chat GPT and this whole um and like just immediately people just just just got into it right away which was a little bit shocking to me to be honest. >> Yeah, chat chat GBT was the most rapidly adopted app in the history of of technology or at least computer technology. Uh although I think there's a lot of resistance coming up now. these a bunch of uh graduation speakers got booed when they tried to to tout how great AI was. So I think >> that's right. >> Um let's come back to some of these questions that are in the local chat because I think this idea of resisting paradigms and and just hopping on to a paradigm shift. I think AI ends up being a good social example that we don't really know and nobody's really I think doing a very precise job of trying to model what the actual change can be at the end when we just do adoption of chat GPT or other LLMs for the workforce. And so, you know, this this idea that we tend to resist a paradigm shift, I think is very helpful in that case because it can slow down something that could end up having at least I'm I'm going to guess in in most cases AI replacing humans is is not a good idea, at least at this stage of of adoption, how young the technology is. Um, and that's something that Cis G said as well as Phil ask talking about those resistances. Um, >> yeah. I was wondering if I could jump in here. >> Yeah. Yes. >> Yeah. Thanks. I I just wanted to say I mean the whole subject of paradigm shifts is is of course extremely exciting and and um it's it's great to be living and also very difficult to be living through such times. Um but we have to be very careful as as Phil mentioned briefly. um there is resistance to paradigm shifts and that resistance is quite reasonable and necessary because you don't want to accept any kind of idea that comes down the pike. You have to be I mean science is supposed to at least for scientific paradigm shifts should be providing the guidance although you know it's difficult to provide the guidance when you're at the forefront of science and uh we can talk about technology paradigm shifts for example as I mentioned in chat there were um the first electric vehicle were uh electric courseless carriages in the 1830s but those were replaced by gasoline powered cars internal combustion engine because gasoline was more basic um oil was found in great abundance and it was cheap to to drill in those days and and and refine and produce gasoline. So that was seemed to be the way to go but that was a major paradigm shift but it was a horrible paradigm shift because of all the problems that's caused us. Um it's caused it certain provide a number of advantages but it's caused a number of disadvantages as well and then we can talk about what's happening in cosmology. You yourself mentioned what's happening in cosmology right now. The accepted model for cosmology is the lambda CDM model which is basically has to do with the uh um the dark energy dark matter. Um and uh the universe uh was born in the big bang 13.8 billion years ago. But observations by Hubble and other telescopes and have have found a number of problems with the with the lambda CDN model. One problem is that um is that galaxies uh don't cluster um as much as they should according to this these models. Another uh problem is figuring out what the Hubble constant is. How rapidly is the universe expanding in the early universe? It seems to be um expanding at a different rate than it is in in the uh in the modern universe and that's not really explained very well by the lambda CDM law. And then there's uh the problem that there seem to be from how from the uh the James Turk Maxwell telescope observations um they found that there are galaxies that are are are seem to be um much more uh much more star formation has occurred than could have happened in the early universe. So is the universe actually older than we think? I mean all these things could overturn the lambda CDM model. We have a completely new paradigm to describe our universe or not. Maybe there just takes a few little tweaks to fix everything. You know, I think uh uh uh uh thank you very much S. I appreciate that comment so much and uh and also you Stephen and you know I think an issue that or I don't know a challenge that astronomy has is that it's observational. it's not kind of experimental and uh for you know I kind of think of that because I'm a big booster of the standard model of particle physics um and uh mainly because it's proven to be so experimentally robust um and has sort of you know uh been able to withstand challenges over many decades uh to it uh not not without some modification of course but I think the basic premise is still very solid. And one of the things that a lot of science communication communicators in the popular uh popular media like to do is to constantly say that some new discovery will overturn the standard model and we're going to have to rewrite science completely and things like that, you know, and uh I'm always extremely skeptical of cla so so I have resistance to uh paradigm shift claims when it comes to like the standard model of particle physics. Um but uh but I do think other fields are more uh are I think um are are more on the foundation is a little bit uh less firm than it is for the standard model. So um so that so I think I think you can sort of res I sort of resist paradigm shifts a little bit selectively that way. >> Yeah, I mean that is that is correct. And you you do have to have some resistance and not just anything. The standard particle model does have its problems. For example, and do we how do you include gravity with the other forces? Not really. >> Good point. That's a good point. >> You need gravitons. For example, if if quantum quantum mechanics is correct, then gravity must be composed of gravitons and gravitons are not included in the standard particle model. Um dark matter particles, what are dark matter matter particles? If they exist, for example, they might not exist. There may be other explanations. So uh there do need to be modifications. Gravity is is it quantized? Maybe it's not quantized. There are other explanations for that as well. So there are a number of things to be considered but for now the standard particle model is is is what we should be using for now anyway. >> Yeah. And I think uh part of the problem there is that uh dark matter and gravity are difficult to um to clarify experimentally, right? Right. I mean, it's hard to devise an experiment to find a graviton or to find dark matter. We're sort of we it's we're very dependent on observations we can make rather than setting up a controlled experiment. So, I think that's part of the challenge. >> Yeah. But I think what I what I also something I mentioned in the original panel too is that one thing we now have is this profession of science which was only starting to really I think appreciate at the time he wrote you know the book in that we have these very formal processes of submitting new claims and new ideas of paradigm shifting ideas. We have people who vet them. we have the sharing of the materials and methods for how we went about and tried to do something. Sometimes we share whole data sets. You see that a lot in high energy or cosmology based data sets. And so we actually now have I think a very interesting new process behind paradigm shifts that uh and that that make it a more reliable thing. Although in some cases you can have influence that slows down progress. But um yeah, high energy particle physics and cosmology are still two great great topics to to discuss. So we're hitting the last few minutes. Um one thing I want to comment on something Phil said is that it's nice to know that we don't know everything yet. Again, that keeps scientists as well as anybody involved in paradigm shifting spaces uh employed and doing interesting stuff. Um you know one thing Gus points out another paradigm shifting idea and this is something you if you look at history this happens multiple times is how do we conduct warfare? Uh you know the idea of tanks being a new technology really changed fair from World War I to World War II as well as having air forces and planes and all that. Uh nuclear power really changed that. Now this idea of drone technology uh especially those that are not uh uh manned or you know piloted by a human. What does that mean you know for new new idea of warfare then fit into also our philosophy and sense of what is the right way to conduct the the noble way to conduct war? Uh that's a very interesting point uh Stephen. Uh you know I do I think it's like I think Ukraine is uh we are in the midst of witnessing a paradigm shift in battlefield tactics that Ukraine is using uh in terms of both the the deployment of drones and in defending themselves against drones. That's all very fascinating. Yeah. So uh so in this last minute any la any comments from the audience? any sort of comments on Baragon's list of AI or social cultural ones? >> Um, well, Sumo mentions this idea of autonomous drones. Um, and I do think that just reminded me that uh uh that uh it was interesting that uh I think was it uh uh we I forget which of the AI companies uh refused to license the Pentagon for autonomous uh basically AI killing machines. Um uh and uh and I do think that uh that in that the Ukraine drones are not autonomous. They're flown by drone operators. Um, and but but I do think that it's interesting that even the AI companies themselves sort of draw the line at Terminator style autonomous killing machines. >> Yeah, I think it comes back to what is the philosophy? What is what is winning a war? What is the proper way to conduct one under the law? >> Yes. Thank you, S. >> Yeah. No, I think anthropic wanted to avoid suddenly becoming liable, legally liable for somebody's death when the US Army used their technology to accidentally kill an innocent and then who's liable for that. Uh, which I think is a big question, big part of the question. So, we're at time. I I what I would like to do is thank everyone for uh attending and interacting with the panelist, the one who did show up, uh, Matt Burr. Stephen, thank you for inviting me today. I really appreciate it and I enjoyed it. Thank you so much. Thank you for coming, Matt.