Google DeepMind CEO on Drug Discovery, Hype, Isomorphic
ฝัง
- เผยแพร่เมื่อ 22 พ.ค. 2024
- Google DeepMind CEO and Co-Founder Demis Hassabis speaks with Bloomberg's Tom Mackenzie about the latest version of the company's AlphaFold AI system (AlphaFold 3), intended to tackle problems in biology including disease protection and treatment. Hassabis also discusses the hype cycle in AI, as well as the timeframe for AI-developed drugs and why he thinks Alphabet subsidiary Isomorphic Labs could be a $100 billion company.
00:00 - Intro
00:32 - AlphaFold 3, understanding dynamic picture of protein interactions
01:28 - Timeline for AI derived drugs. Partnerships with Eli Lilly and Novartis
02:03 - Isomorphic Labs, drug discovery
02:05 - Potential for mRNA vaccines
03:25 - Competitive threat from OpenAI
04:24 - Sustainability of AI
06:10 - Rationalization in AI sector, more closures and failures
07:14 - UK AI safety summit
08:47 - Big election year - are Labour as invested in AI?
09:20 - What are the big things coming this year?
10:31 - Isomorphic Labs, under pressure to commercialize?
12:20 - What keeps you up at night?
--------
More on Bloomberg Television and Markets
Like this video? Subscribe and turn on notifications so you don't miss any videos from Bloomberg Markets & Finance: tinyurl.com/ysu5b8a9
Visit www.bloomberg.com for business news & analysis, up-to-the-minute market data, features, profiles and more.
Connect with Bloomberg Television on:
X: / bloombergtv
Facebook: / bloombergtelevision
Instagram: / bloombergtv
Connect with Bloomberg Business on:
X: / business
Facebook: / bloombergbusiness
Instagram: / bloombergbusiness
TikTok: www.tiktok.com/@bloombergbusi...
Reddit: / bloomberg
LinkedIn: / bloomberg-news
More from Bloomberg:
Bloomberg Radio: / bloombergradio
Bloomberg Surveillance: / bsurveillance
Bloomberg Politics: / bpolitics
Bloomberg Originals: / bbgoriginals
Watch more on TH-cam:
Bloomberg Technology: / @bloombergtechnology
Bloomberg Originals: / @business
Bloomberg Quicktake: / @bloombergquicktake
Bloomberg Espanol: / @bloomberg_espanol
Bloomberg Podcasts: / @bloombergpodcasts
Demis Hassabis deserves Nobel Prize in medicine for the invention of the AlphaFold system!
Agreed, 💯.
A great gift to humanity. Alphafold 2 was cited in some 1,500 papers, this will go down the same or more. We truly are heading to the end of disease and birth defects. There are currently 1,500 genetic treatments in various stages of testing many of which will come to market.
Absolutely he does!! A true visionary
@@brianmi40
I hope so, but there always are some obscure reasons, that you won’t get to 100% eliminating defects.
@@antman7673 Keep in mind that as we check off each problem and the pool of remaining medical problems shrinks, the ratio of people/AI/effort focusing on that smaller number grows. We are mastering biology, there's no magic and powered by AI nothing will not be revealed over time.
The interviewer seemed well prepared. Good interview
Extremely well prepared and very articulate.
to me, he sounded like these questions were prepared for him by someone who only cares about revenue and imvesting.
the interviewer himself, on the other hand, seems like he can't wait to wrap things up.
interviewee seemed like he had so much more he wanted to talk about. including his enthusiasm and message that comes with alphafold3.
buuuut of course we gotta think about the people with ADHD too.
Tom Mackenzie comes to this interview with the most penetrating question set I've seen over the last year, and Dmis Hassabis ably responds. This is a valuable interview.
One thing that stands out with Demis is that his drive for success was never about to 'outcompete' others - he aims to find out the truth and help push research and innovation to get there, for everyone, and not just the few.
That's what they said about Mao. And Elon the latest incarnation of Science fiction on the taxpayer life savings. Do you serious believe humanity will 'live among the stars'? He's 'just doing G-d's work.'
Maybe I'm jaded. I worked on two Star Wars projects. It became obvious the sci-fi hype was used to siphon Mom&Pop 401ks for The Right Stuff and the Bankers.
Not a single Star Wars project ever deployed. The last 'major biochemical breakthrough' genius is now doing prison time.
dam the universe did good to make someone like demis
Add Shane Legg to the list as well. Amazing human
I have a degree in biochemistry back from 1992... and we actually talked about this in some classes. Imagine a steel factory where you need to precisely bend a one mile long rod of metal 10s of thousands of times and put each bend at a precise position. Actually figuring out where each bend needs to be is a huge task all on it's own, but then figuring out the factory configuration to where you can do all of those bends, while still allowing the rest of the rod to fit move through the factory is a Herculean task. We knew back then that someday computers would not only be able to design the drug, but also be able to describe the process we needed to use to make that drug.
So then a steel factory comes along that bends every long rod of metal precisely tens of thousands of times all at once?
@@squamish4244 you cant synthesis organic molecules that way. There are hundreds of steps that need to happen in a particular sequence.
@@PeterSedesse Just trying to find the right description of what AlphaFold 3 has achieved using your analogy.
@@squamish4244 that really is the power of AI.. I don't think the human brain can envision that process on the scale of an organic molecule that complex. It is like a chess player who can see mate 9 moves ahead, but this would be tens of thousands of moves.
@@PeterSedesse Holy shit. 99% of the population has no idea this is happening. The chatbots get all the attention.
Demis is in another league, a genius, the real AI GOAT
Unlike sales man Sam this guy is a real hands on guy
Thank you demis !
Ai is already used extensively in the most advanced chip designs as Jensen Huang stated for Nvidia. It's only natural that we see it play a major role in drug design, this is going to be an incredible next decade.
AI is not used in the chips, but rather optimization algorithms. They have been used in the industry for years. If what you say is true, which it is not, AI has been used since the 70s to improve industrial processes. Algorithms are not necessarily AI, but AI is algorithms.
He's my fav AI oddball. Like, if you dropped a tenner and he saw I reckon he'd pick it up and notify you.
He went to my highschool. And back then he was already a genius, made millions out of video games as a teenager
Demis a genius
This is kinda important, world.
This is what I like about Google's Deepmind. They still have their super cool and humble boss in place. Not fighting like the OpenAI folks.
Great to see him involved in drug discovery, we are getting ever more closer to age reversal, a great future of possibilities exist!
As a computational biologist from a small biotech company, I am scared that these bigtech AI companies will overtake my job. But as a human being, I hope they make a huge progress to find better drugs in more efficient way.
I am also thankful that Demis is interested in biology and various science fields rather than just making targeted advertisement.
I hope the AI development will have a huge impact on the modeling support of drug discovery and development helping out in increasing the efficency. Still the elephant inthe room is translation or lack of understanding thereof, meaning generation of buckloads of biological data for you to process.😊 We need to remember how the data behind alphafold was generated, how many PhDs and industrial scientists are behind the PDB database? That was the hardest work enabling alphafold. That doesnt off course cast any shadow on the clever work creating alphafold. It is a big step.
Great work! More spotlight should be placed on such advancement and research. A clear economic model should emerge on its application with the right measures to prevent its misuse.
Great interviewer! Succinct questions. Covered a lot in 12 mins.
Most of the big TH-cam podcasters would have taken 3 hours to get the same info.
He should be Alphabet CEO.
Demis Google CEO and sander alphabet CEO
Here is something that Alpha fold or isomorphic labs can do. Suppose you have a molecule consisting of a specific set of atoms. Perhaps it is not catering to the specific action you want it to do. You can always replace it with an atom that is of the same outer orbital structure, like say replacing sodium with potassium or carbon with silicon and see how the molecule changes and whether it now caters to the job at hand. Just sayin.
The way the AI could figure out what happens if you replace sodium with potassium is train on already available molecules that have sodium replaced by potassium and learn what happens when you do such a thing. Or maybe even manufacture some of the molecules and train on them.
Master class in avoiding the words "compete" or "OpenAI" in his answer, huh.
The people saying that Google are left behind in AI are lunatics
I'd love if demis and sundar just had like a duo interview together on a podcast, for like 90 minutes
Demis Hassabis, the next Alphabet CEO.
Amen
Hopefully - also replace some of the current management at Google with his DeepMind peers
No, he’s an Engineer not an Administrator.
The current Google CEO is an engineer. All previous Google CEOs were engineers.
@@StatMachLearn Whatever he was, he is an administrator now.
Can ML predict the minimum energy/lowest entropy for proteins?
Turn down the sound and watch the gesticulation
hah ya
Many oddball AI people are similar haha
AlphaFold 3 is such an important project. Wonderfull
By end of the next year we will probably have a virtual cell - and Demis should get a Nobel Prize for all his contributions (and become next CEO of Google)
Game changer in precision tx designed drugs
Google should start a bio company or a branch in google like google cloud google biotech
and i think this could save googles future
Google will be able to patent their discoveries thru Isomorphic.
I am watching a future Nobel Prize winner.
9:48 health AI Agents 😮
the fact that people invest without even knowing the possibilities i hope it leads to a chance to invest in AI at a cheaper price because of less knowing people not believing in AI due to it underwhelming in the short term. im 100 bullish on AI long short term but especially long term
Buy Google stock now while it's less than $200
00:00:04 Alpha Fold three predicts biomolecule structures.
00:00:25 Understanding protein interactions for drug discovery.
00:02:50 Potential implications for RNA vaccines and biologics.
00:03:32 AI advancements equip Alphabet for competitive challenges.
00:04:44 Sustainability challenges outweighed by AI benefits.
00:06:28 Rationalization expected in the AI industry.
00:07:18 UK's pace in AI development and infrastructure.
00:08:29 AI's potential benefits and responsibilities for society.
00:10:45 Revolutionizing drug discovery for societal and commercial value.
00:11:39 AI-generated drugs may impact drug approval processes.
Would Hassabis please address Human Microbiota Project? It's the low hanging fruit to youthfulness and longevity. The scientific evidence is "The Longevity Paradox..." by scientist and heart surgeon Gundry and the physical demonstration is Jhoon Rhee, father of American martial arts who was like a teenager at age 83. You don't need synthetic drugs because your microbiome can produce what you need for youthfulness and longevity. Hassabis should solve the human microbiome project at the protein levels.
Can't wait for the drug that will get me a full head of hair!
One word: Wig
rheumathoid arthritis, goat, vih, herpes, aging, genetic anomalies, cancer, diabetes those are the big shots that I can imagine right now
Hairloss
@@LarsRyeJeppesen Are you being serious? That's the least pressing health problem.
@@fupopanda add a few extra inches to your height?
We could decrease everyone's height and use less resources 😂
@@fupopanda tell that to the 25% of women also affected. Is it an either or? Did not know this
Chapters (Powered by ChapterMe) -
00:00 - Alpha Fold 3 Dynamic protein folding, implications for drug discovery
01:17 - Alpha Fold 3 predicts protein binding and drug discovery
01:39 - Generative AI drug development partnerships with Eli Lilly and Novartis
02:36 - Alpha Fold 3 opens up new vaccine opportunities
04:36 - Challenges and benefits of generative AI
07:41 - UK planning laws hinder AI growth
08:07 - Economic and political opportunities for AI in UK
10:16 - Bringing gaming and drug discovery back to Is Labs
10:45 - AIs potential in drug discovery and approvals
12:01 - AIpowered drugs could be beneficial for patients in future
12:33 - AIs diverse nature
The Her-like chatbot gets all the publicity, of course, but this is a WAY bigger deal. I guess this is a really hard thing to be alarmist about. "Oh no, we'll have way better medicine and cures for all sorts of awful diseases of the body and brain! That's...uh...umm...sh*t. We can't turn this into a frightening headline. Forget it."
I read it as Demis Hassbis discovers drugs
Next couple of years 🤔🤔
Demis Bot 2.0 required to handle these interviews while Demis OG focuses on his science.
AI Horsepower: Canny Watt coined "horsepower" 1806, so mill and factory workers could calculate savings of replacing horses with newfangled engine. How many Ph.D. protein researchers did Alpha fold replace, how money did it save and what are those Ph D protein researchers doing now cuz you, Hassabis, are the new Watt and your AI has lots of horsepower. So, 29,998 official NOC horses/vocations are next?
Successful vaccines? Haha, my health has been noticeably altered since the vaccine, I got it as a hospital worker very early on.
Man I feel you. I was forced to be vaccined 5 times, and now my tsh, t4 is all messed up
It’s a please to hear Demis Hassabis. Legend in the making 🙌🏼
AI-assisted drug development and soon enough...hopefully better quality of life
10 year lifespan extension, minimum, by 2030 due to cellular reprogramming.
Y not grow graphene intead of diamond?
Roadblock? Money & time, always. Gotta figure out how to pay people for all their data…even if it’s generated data, the seeds have to come from somewhere.
More kids need to study biology
💚
If i have to bet humanity on either Sam or Demis, I would pick Demis. Less hype, more science. Scanning eyeballs is not science.
Sam seems like more of a "hype" guy. A business guy. The front facing guy, but does he really know 1/100th of what Ilya knows?
All I saw was "Google DeepMind CEO on Drug....."
This guy's the brains behind the whole thing?..
a lot of hand gestures going on with these two, hand jive masters
Bloomberg you can use AI tech to remove the static noise before uploading the clip. How naive to not do that and so ironic for this being a video about AI and its applications.
I Hope alpha-fold 3 helps in cancer and.. drug discovery soon.
AI in pharma is great but it just another tool for these Pharma companies to create and patent drugs iust to sell them at exorbitant amounts. The layman will still be at the mercy of their Insurance providers.
So when is my car driving me everywhere, 2040?
Oh great, as AI unlocks biology, big pharm has the keys 😢
Go 4 the perfect drug...
This is like the beginning of the movie "I Am Legend", zombie apocalypse incoming 🧟♂️
Hmmm, Nobel Prize incoming!
It can be EITHER of huge benefit to society OR a hundred billion dollar industry not both, a huge benefit to the part of society that can afford it is a slogan that doesn't exactly roll off the tounge though.
Given the corporate climate we live in I just don't believe what these Ai guys say no matter how mild mannered and virtuous they present themselves as on camera.
Yes it will be good for the climate but first we need to burn everything we can find to power our data centers.....Now where have I heard that argument before, Ai is going to make current inequality look like a fairy tale.
The cures might start off expensive, but they'll eventually trickle down to the poors
Great timing, much better interview than Sundar's
Surely He’ll be replaced in 2 years by an AI robot at Google .
Google is ideologically stuck. Scarry for them to have this power under the evil of their views.
Isomorphic , I misread that as Islamophobic
I find it interesting that he does not mention Reinforcement Learning (RL).
It was revealed in a 2018 email that came out of the lawsuit between Open AI and Elon Musk, and probably hence, why Google stated last year that "we have no moat". RL is most likely describing the "core technology...from the 90's" in the email (though RL was not explicitly mentioned in the email, it's a big hint),
RL was funded by the USAF (i.e. US tax payers) prior to 1997, and there is an actual "Q*" described in the second edition of Reinforcement Learning (RL) book by Sutton and Barto (even though it is lower case q*).
The 2nd edition of RL book does not mention K-means clustering, but this is used in a 1997 thesis (that almost no one knows about) that also uses Fuzzy Logic (a merging of mathematics and language, for state space classification and control of an inverted pendulum, not the exact title, but it is at ASU....maybe) applied with RL.
interviewer needs to relax
Coke bro
Pretty sure the host did a bunch of meth to try and keep up with the guests brain power
What does Deepmind think about Google ditching all kinds of half baked projects 😂 they can’t even keep the same damn name changing it from Bard to Gemini 🤦🏼♂️
Is there substance to Al Jazeera's accusations against Alphabet?
HYPE! 😂
MOST RESEARCH IS TRIAL AND ERROR... NOW WE HAVE ONE MORE INTRUSION.. AI... AFTER AI GIVES OUTPUT, IT WOULD BE AGAIN TRIAL AND ERROR.. AFTER FINDING IT WON'T WORK AFTER SOME 100 TRIALS, WE WILL GO BACK TO PRE-AI STATE AND ADJUST PARAMETERS AND RUN AI AGAIN, AFTER WHICH THE PROCEDURE IS SAME... HOPE IT REDUCES TRIAL AND ERROR EXPERIMENTS AT LEAST BY 50%.. THAT WOULD STILL MAKE AI VERY VERY USEFUL
Accurate, robust simulation helps reduce trial and error so that when the weights are transferred to real world application, there can be less trial and error, but nothing is going to be perfect, but RL (Reinforcement Learning) can come pretty darn close, perhaps closer than any human can.
Why are you shouting? Never trust people who shout
You forgot to take your med?
So are we going to return to the drawing board after using CAD systems? I don’t think so.
2 years? Too slow. You guys will get left behind for sure. Things are gonna move at a much faster pace. Try to keep up!
Completely false. Things move slower and faster than people usually believe. That is, for people who believe that all this will happen quickly, it will happen slower than their minds imagine. Those who think it will go slow, things will happen much faster. Neither one nor the other is right. The truth is that things will go at the pace set by events.
Demis is probably deliberately giving a longer timeline to be conservative. He's not the Musk, hype type
@@raul36 It's all relative i supose isnt it? faster... slower... that's all subjective. Quickly for me may be slow for you.
@@Psmitty97 It's possible. But Musk has achieved a lot more than Demis overall and shown the world possibilities that Demis never brought to the table. Not saying that Demis hasnt made significant contributions. He has. I just dont see him on the same level as someone like Musk. And if others were right in their predictions about Musk, he should have failed a long time ago. But he didnt! That's a good thing :)
this won't age well...
Why not. Alpha Fold 2 is already a proven technology. It predicted all 200 million known proteins (folding).
shit
Funny to see: A.I. viewed trough the lens of how much money can be made with it. Like a bunch of chimps in a supermarket.
I love this kind of interview. Inteeviewer says nothing and the guest tries to advertise. I learnt nothing too. I guess its free audio
Imma stick with Google EVEN THO Microsoft is the leader.
Microsoft is the leader on paper. Goog's been all-in on AI since 2011. Way before AI was ever a twinkle in Satya Nadella's eyes.
Meanwhile i need 10 tries to get Google Assistant to play a song ... yeah i don't buy that google is the AI company
Think about all those doctors that will lose their jobs.
Comparing this guy to say Elon Musk on answering technical questions, he is just using buzz words and does not actually say anything that a well informed outsider would say about AI and drug discovery. As someone in the business he said nothing of value.
Do you even know who he is? The guy is an absolute genius, and you’re treating him like he’s just some random idiot lmao. His achievements aren’t as impressive as Musk’s (whose are though?), but he’s almost definitely more intelligent.
feels scripted, not a real interview
Why is google repackaging Alpha fold? Will they also release alpha GO 3 ? Google is just reaching at this point
yeah rushing AI agents ....what could possible go wrong