> Our general approach is to automate the experimental loop. We think this approach is broadly applicable across many different fields of science and engineering. We’ll initially focus on ML research and engineering, but believe the approach can help with important subproblems in nearly every one of the fourteen <at>NAE Grand Challenge problems. We think doing this well requires strong expertise in machine learning as well as large-scale systems.
They make many bold promises, but their core goal is neatly encapsulated on the website:
"Imagine a future where a handful of people can conduct scientific research and engineering tasks much more rapidly, and with higher quality, than massive teams of scientists and engineers do today."
This is the same goal as every other AI company out there. Automate away the human employees and let a small number of "people" (note that they do not say scientists or engineers for this part) take the credit and financial rewards for every good thing this human-free system produces.
I love this question. If history is a guide, some version of that is what will happen (employment doesn't seem to go down, despite technological progress and population + economic growth).
I suspect the root cause is that it's harder and scarier to imagine good outcomes. It exposes us to disappointment, and when you do it publicly, it looks "crazy".
Another explanation is that there's no direct consumer for "more." Individuals, corporations and states are not in themselves interested in "a larger amount of science," or anything analogous, despite the fact that they would all benefit ambiently.
The net effect is that it's only "safe" to claim reduced risk (i.e. lower costs).
> Similarily, everyone becoming their own accountant / lawyer / other professional services.
These professional services will defend themselves with gatekeeping. Suddenly it doesn't depend anymore on the quality of your legal advice, but whether it has been stamped by a qualified Lawyer. It doesn't matter that your taxes are correct, but whether they are submitted by an approved accountant. Etc.
They already do protect themselves. It’s what licensure is for in large part. I’m not saying licensure doesn’t have other benefits, many are plainly obvious. But licensure is also used as plain old gatekeepong.
Because one of the primary goals of using AI for companies is to pay as few humans as possible for the same amount or more work, or preferably pay no humans at all, because that makes them more profit.
What happens is that your competition uses AI, your investors factor AI in, and your customers use AI agents to find what they need and can switch providers with greater ease in the agentic era. So even if a company does nothing the economic environment around it changed. The AI boost is mandatory but produces few winners, it's mostly a scramble to swim harder just to stay in place. Benefits are competed away fast.
So it does not follow that companies can bank the savings from firing people. Anything AI can do for me it can do for my competition as well, and humans still make the difference. The big question in the AI age is "why pick me?" why hire me, why invest in my company, why buy my product, in a sea of similar products made by everyone. A differentiation crisis accentuated by AI.
Also, wouldn't anyone with half a brain use the human-free system to produce another human-free system that was no longer controlled by the "small number of 'people'"?
Why do you think you'd be given access and permission to do this? If a company genuinely cracks this human free system problem, why would they open it up, instead of simply outcompeting everyone that doesn't have their product?
In science fiction, the AI agent has written the training loop management software and included a back door to prevent that from happening (or found a way to talk to the training loop management agent and convinced them to not listen to the evil human when it tries to do brain surgery on the AI agent.
Also, when the human reaches for the power switch the AI agent uses a flaw in the power management software to weld the switch shut with a big power surge, killing the human with a huge electric arc in the process.
I don’t think whether we will get there, but the stories of LLMs escaping their sandbox make me think we’re moving in that direction.
Why wouldn't it? There's a lot of research going into AI alignment, which means finding the best way to train an that AI isn't going to get in the way of making the people training it as wealthy as possible.
I don’t think we have the data to disprove the stronger claim “societies collapse”, and I don’t think being unfair (whatever that means) makes societies collapse earlier. Did slavery hasten the fall of Rome, for example, or the Gulag the fall of the USSR?
As to “whatever that means”, I think that’s hard, if not impossible, to define objectively. Catholic dogma says the Pope is the representative of god, for example, so catholics (less so in modern times, I think) don’t question his decisions. Many would call that unfair, even if the pope would be elected 100% by merit.
I mean of course the universe has no need to be fair - that's a frankly asinine observation in the context of a discussion about social policy and resource distribution.
the universe doesn't require opposition to slavery either but I'll be bold and assume you oppose it anyway
My best research is one that is seeking funding but none is forthing and by sheer guts and whatnot makes a breakthrough that basically funds itself. No investors. No funders. Definitely not public.
Because of the negative externalities. It’s the same reason you shouldn’t do any other thing that personally benefits you but imposes a greater cost on everyone.
My main point is that people shouldn't just swallow the noble and lofty sounding PR. These guys are the same as everyone else in the sector. Don't ignore the harmful or scummy things they'll inevitably do. Hold them accountable. If they are as noble as they sound, they should agree with me.
> "Imagine a future where a handful of people can conduct scientific research and engineering tasks much more rapidly, and with higher quality, than massive teams of scientists and engineers do today."
I don't think the US have this capability because you guys don't really have manufacturing that is really necessary for scientific research.
For example, if I want a highly toxic chemical, how difficult it would be to procure that in the US vs China?
There is but the cost is way, way higher from what I can tell. If I want a customized metal shielding with awkward shapes, I can just walk down to Shenzhen and have 10 people clamoring to make it. Does the US have the same?
Maybe not 10, but my company (in the US) has relationships with several machine shops nearby (some within walking distance of our office) that do custom metal fabrication for us, and sendcutsend and xometry can handle more exotic stuff without having to ship it from Shenzhen.
I think the bigger advantage of Shenzen is more for actually manufacturing at production volumes cheaply, although maybe if you need something really exotic? Custom metal fabrication is not a good example here, I have a neighbor who runs a custom CNC parts business out of his garage (said garage is largely taken up by his CNC mill).
If time is of the essence, or rapid prototyping is needed, the shipping hurdle is a real problem. Going down the street for the goods vs having them go across the globe is quite a difference…
Are you a scientist? People have idealized views of science.
>>> Yes but science is well-structures and practically designed about repeatability so its a lot easier to automate than "softer" disciplines.
What AI is up against is that science is already automated to a high degree, so the AI doesn't just need to automate things, but it has to automate things better. Also, a lot of science work is in dealing with boundary conditions, edge cases, exceptions, hypotheses, and so forth. That work is essentially chaotic.
Do I think AI can improve automation? Sure. Everything I do in the lab is automated, and I use the AI coding assistant.
Yes I am, which is why I said that, its much easier to automate over the chaos with a coding agent than with a script. They are still far from doing anything productive on their own but its easy to imagine designing a search space and "letting them go", especially since coding agents are approaching the point of being able to reproduce papers reliably.
If you think the public rejection of AI is weird, wait till you catch a whiff of the marketing!
The future described by these AI labs isn’t like previous waves of technological progress. With those, technology displaced some/many occupations, but it left open the door to other, higher-valued career paths. What these labs are proposing to do is to dissolve virtually every path to upward mobility that exists, simultaneously. Even AI research itself would seemingly require nothing but a checkbook.
Mind you, I think it’s a load of hot garbage. I don’t see the evidence that LLMs are en route to the future these labs keep promising. But it is a dark and ugly future that they claim to be racing toward, for reasons.
Humans have proven to be incapable of making real progress in science given the effort. Especially academia. We haven't cured many diseases and cancer when we should have (and no im not an amateur and yes I do believe we can "cure cancer". Debate me).
We already have a trillionaire, whats the difference? The only difference I see is that people who were previously rich, but considered themselves middle class, are now realizing they are actually poor just like the several billion humans around the worldanyway.
Many cancers are "cured" in the sense that they're detectable and treatable. My grandmother lived 35 years longer than she would have by detecting breast cancer early and getting a mastectomy.
There is currently a $800 SOTA blood test that can detect most cancers before any symptoms. Maybe a decade until it's a routine part of your annual blood test?
Whole genome sequencing costed $2.7 billion in 2003. You can get it done today using a mailed kit for $400.
HIV went from death sentence to all-but-cured in 50 years.
> The solution to most of these problems lies in policy, not in new tech advancements.
That is not mutually exclusive. If technological advances result in a given technology becoming cheaper, more scalable, and easier to deploy, they also make it easier to advocate for and implement the relevant policies.
You can think of it like this: our "political technology" is not good enough to use solar energy at its current prices to replace fossil fuels as fast as we would like. Well, what about if we cut the price of solar by a factor of five? Perhaps it will be good enough then.
I don't think that's the way I'd describe it though. It's more like "our political technology is not good enough to reliably do good things instead of bad things". Working around it by lowering prices of this or that is like if you're bad at darts so you just make the board bigger and bigger. Okay, maybe you'll hit it more often, but you didn't fix the problem that you're bad at darts, so everyone still at risk of having their eye put out by an errant throw. Moreover, I don't think I'm the only who finds it a bit perverse to accommodate to failures in that way rather than trying to actually make things better.
It can indeed feel perverse, but we have to do what is realistic, not what is unrealistic, especially with climate change where we are on a clock. Fixing our politics is probably much, much harder than lowering the price of solar to such low values that the market would simply have no choice but to adopt it, or developing other technological solutions.
> If technological advances result in a given technology becoming cheaper, more scalable, and easier to deploy, they also make it easier to advocate for and implement the relevant policies.
That sounds like just playing further and further into the game of the corrupt leaders. Who do you think would profit off that 5x margin? Would that margin come more likely from a scientific breakthrough of via some new exploitation of natural resources or human labor?
It's yearsss past time that our leaders should have changed policy.
Making progress requires first not dismissing your ideological outgroup along such lines, and instead trying to understand what actually motivates them.
> I do not see how the second sentence follows from the first.
Point 1 on the list is "Make Solar Energy Economical".
Solar is economical today - mostly due to China investing (and heavily subsidising) in solar for the past couple of decades; but compare to the US where certain particular big-businesses (oil companies, mostly) were instead cynically funding disinformation efforts and getting into bed with the Republican party (which dovetailed with the GOP's allying with other science-denying movements of the Bush Jr era like creationism and public-health matters with abstinence-only sex-ed and defunding gun safety research efforts) - means we're decades behind where we could have been...
Consider an alternative past, where the GOP had the backbone to resist the oil industry's corruptive influence and instead made a big bet on American Solar; it's entirely possible that instead of MAGA today we'd instead have a right-wing coalition strongly supporting solar and wind energy because they align nicely with American rugged individualism - whereas the current situation on the right is an unprincipled farce with inconsistencies in policy positions at every turn.
This is not a particularly new struggle: Jimmy Carter installed American-made solar water panels on the White House in 1979, then Reagan tore them out.
It's a lot more complicated than that -- Solar is great for discretionary, opportunistic, and time-shiftable additive demand, but not for replacing all-day baseline load. Using solar to dip into baseline demand sometimes is catastrophic, because the "profit" is exploiting its failure to provide consistent, reliable power but charging "default" prices that have an implicy assumption that power is consistent and reliable.
Nobody suggests moving to 100% solar. Some off-grid homes do it, but generally there are multiple complementary power sources available. There is no assumption that any single power source always provides consistent output and it wouldn't make sense, because production needs to match consumption, which is variable.
There is also no default price on energy markets, it fluctuates with supply and demand. Dynamic pricing by itself is enough of a reason for industrial users to build up their own power storage, which allows them to time-shift consumption from the grid.
Time-shifting is definitely going to increase, but it's not a bad thing. Look at how battery storage has made electricity cheaper and more reliable in California.
Start with separated sewage/wastewater and stormwater drains. Then accredited and highly scrutinised wastewater treatment and discharge into water bodies (or see below for a high-tech solution). As for clean water to the home, direct those stormwater drains to new reservoirs which sustain freshwater aquatic life. Protect aquifers from over-drainage, and build pipelines from water-abundant regions to water-scarce regions.
To reclaim waste water or treat unknown water sources back to potable/semiconductor standards we have ultrafiltration, reverse osmosis, UV treatment, pH adjustment, fluoridation, desalination, softening (which is generally obviated by RO...). This is basically Singapore's NEWater.
Good sanitation is a financial and political problem. The engineering has been solved for decades now.
Probably want to jump 1 generation ahead directly compared to captive Chinese investment with fully automated US factories churning out Silcon perovskite tandem panels directly with few inputs and electricity
People on HN keep saying it is, but I'm still not seeing it. Companies are building datacenters in vast, sun-blasted deserts and still choosing to power those with natural gas. This in turn makes people complain about emissions pledges being reversed, but if it were economical, no pledge would be needed.
Of course, USA has cheaper oil/gas than other countries. But if you look elsewhere, rich countries are subsidizing solar, poor ones are basically not using it.
In the chart for section 3, how many countries have seen their share of electricity being generated by fossil fuels increase in the last 5 years? Only Canada.
Take a look at the charts for Pakistan, Australia, Nigeria, and China for the last few years. Pretty dramatic drops for fossil fuels generation.
They cherrypicked 15 countries. And still, some of those still had renewables decrease since 2000 like Nigeria, others saw an increase but it's still way less than fossil, and others like China are heavily subsidizing solar. I don't doubt that it's economical for individuals when the govt is subsidizing it.
The "world" chart shows an increase from 19% renewal to 34%. Did they cherry-pick that? (Also, "European Union" is more than one country.)
> I don't doubt that it's economical for individuals when the govt is subsidizing it.
Does that distinguish renewables from fossil fuels? Haven't governments been essentially subsidizing fossil fuels (not least by allowing environmental externalities to be ignored) for as long as they've been in use?
Because a lot of countries in the world, especially the EU, are subsidizing solar. That doesn't make it economical.
Externalities ignored from fossil fuels, yes. That's not a subsidy though. I'm not saying they should ignore it, but if they do, they aren't the ones who pay for it.
Look at Australia then. Millions of homes already using solar yo basically power their homes for free most of the time. Yes it was subsidized, like oil was and still is. Solar without subsidies is already miles better than oil and gas.
Australia is a rich country that subsidizes solar, and they're still 90% fossil according to their Wikipedia article, so idk why the mismatch with this article.
Solar is dirt cheap in China, where 85% of panels are produced. The problem is they're made in China and face tariffs/import bans in the US/Europe.
Nat gas is preferred for AI DCs because it has faster time-to-market, doesn't have the intermittency issues. Training on solar + storage is an issue because of network synchronization.
I get the gas turbine for semi-temp power when there's not enough grid support, but Google leadership is talking about doing this long-term and at scale: https://www.gstatic.com/marketing-cms/79/80/fb229abf40efa81e... . Not a single mention of "solar" or "renewable" in there. Are they just trying to appease Trump administration?
Google may not have written about it in that document, but they're funding what will be the highest storage capacity battery system in the world for a renewable-powered data center in Minnesota:
Data centers need lots of power 24/7 and regardless of cloud cover. Solar is great to reduce your daytime bills but you still need other methods to cover the downtime.
I would be surprised if data centers didn't put in gas _and_ solar.
They might be, but if they are, it's because a new administration might subsidize solar again. Meanwhile Trump seems like he'd be against solar even if it were economical, I guess cause oil companies.
What makes me pessimistic is even during Biden's administration, these companies made meaningless pledges more than actual changes. This suggests that the most profitable thing to them is fossil.
Many of these problems don't seem scientific at all, but rather a problem of political will.
As you said, Solar power is incredibly economical. There are plenty of ideas around putting them over farms, or parking lots en-masse to provide cleaner energy.
Access to clean drinking water, while certainly scientific in some situations, is also a problem of political will and money.
Restore and Improve Urban Infrastructure - It's infrastructure week!
"Solar PV with storage = solar PV installation paired with four-hour duration battery storage, scaled to 20% of the output capacity of the solar PV."
Sometime the sun goes away for more than 4hrs.
That may be OK for closed-ended systems (turn off the science at night and during storms), but not for open-ended systems with diverse user demand.
4hour batteries are competitive with gas peakers to match high demand during and after sunny times, but solar needs gas peakers or similar to over for non-sunny times.
It is not economical compared to alternatives that is why you have to have government to force people to do things. In many places such as Pakistan where solar power is not economical on paper is actually very successful in practice because it is actually profitable.
To make solar power practical and economical you need may a square foot of solar panel being able to get enough energy to power and entire home for a week
Not sure what you’re talking about here. We can’t replace all energy needs with solar but it’s clearly one of the cheapest energy sources and with the added benefit of low capital expense to get started so you can set it up in distributed grids without the massive expenditure to support nuclear installations.
Very honorable effort, but a lot of these seem to touch heavily regulated industries impeded by unwise or outdated policy no less than by the lack of clever engineering - medicine, education, urbanism, energy. I wonder if they've given some thought to the key blocking factor as well.
Why not just add 'mind control' while you're at it.
Only death stops stagnation in the end. Without death, especially if death can be avoided by the rich and powerful but not the poor, life will get much, much worse for the average person (until only the rich and their automated capital remain I suppose, in which scenario they will simply turn on each other).
When one of them dies, they want to leave behind a puzzle so complex that entire groups of the population dedicate their lives to solving it within the virtual world. They look old enough to have a lot of favorite 1980’s and 90’s pop culture references, so those will probably be the clues.
I'm guessing this might be about "teleoperation" (like remote surgery via robots + VR) and being able to remote training as well. The binocular vision VR gives you compared to flat screens help a lot with depth perception for precision of incisions for example.
1. Make Solar Energy Economical - Disallow fossil fuels
2. Provide Energy from Fusion - See 1
3. Develop Carbon Sequestration Methods See 1
4. Manage the Nitrogen Cycle - See 1
5. Provide Access to Clean Water - See 1
6. Restore and Improve Urban Infrastructure - See 1
7. Advance Health Informatics - See 1
8. Engineer Better Medicines - See 1
9. Reverse Engineer the Brain - See 1
10. Prevent Nuclear Terror - See 1
11. Secure Cyberspace - See 1
12. Enhance Virtual Reality - See 1
13. Advance Personalized Learning - See 1
14. Engineer the Tools of Scientific Discovery - See 1
FF is the real threat in time, money, health. Can't sweep aside that it will destroy most life on Earth and we'll never get to the other things if we are at the mercy of FF
To understand, same reason you reverse engineer anything. Doesn't have to have a further goal than that, understanding the brain better helps in so many ways. But like most technology, obviously can be used for bad too. Should we just skip researching some topics then?
I'm all for alleviating psychiatric/mental health disorders, but yes, some topics are worth skipping research on. For example chemical/biological/nuclear weapons, human cloning, and unethical gene modification.
I just like to challenge myself, as an engineer, with the idea that not everything has to be engineered and optimized. What if we simply left some things unexplored and mysterious, and trusted nature and our own human capabilities?
A better way of alleviating psychiatric/mental health disorders might just be to focus on societal factors.
We have a good understanding of the function (and more importantly dysfunction of) kidneys, lungs, heart, etc. from high level to cellular level.
For brain, our understanding is fuzzy, more like "this part is important for that behavior" or "here is how neuron works" but we don't have a holistic understanding.
If we had that, we could more easily diagnose and treat neurological disorder.
This list smells of so much tech bro “I can do better than the people that have been working in the field for 20 years” egotistical attitude that permeates silicon valley. Why do software engineers think they are smarter than everyone else? Is it because they earn more than most people? But then bankers and finance bros should think they are gods?
Also, solar energy is already economical!? Do they mean more economical?
>This list smells of so much tech bro “I can do better than the people that have been working in the field for 20 years” egotistical attitude that permeates silicon valley.
Well, they did solve some math conjectures recently that the people working in the field for many years did not... Also AlphaFold.
Treating autism (and preventing it) is not eugenics. That's a talking point some folks raise, but we're talking about intensely pervasive autism, not neurodivergence. Really, please try to do better in your arguments than immediately saying your opponent is like Hitler.
Sometimes you got to find a way to buy the silence of your top employee, to prevent them from going to the competition. This "start-up" is shallow as hell
Given the resources that would be available to them at Google: compute resources, data, etc. It's clear they want absolute freedom. Good for them. At this revolutionary turning point in history, I want the smartest people working in whatever area they want.
This reads as hype of the kind: "well, we can't reliably do these rather mundane things with AI, but we're going to run it extra hard and extra long in some novel way, and it will do amazing things".
The issues are with verification and with detecting drift from the goal. These are related, if not roughly the same issue. And, if they can solve this, then they will have essentially fixed AI. Maybe even AGI.
But, if this were the goal, then it seems more reasonable to solve the relatively more mundane verifiable challenges (e.g. generating solid, reliable code). Then, working up from there.
And, that's exactly what gives this the hype smell. No use for solving problems that don't get the oohs and aahs. Just straight to NAE Grand Challenge problems.
I think people are missing what this really is: Google giving some of its most senior engineers the best retirement home to keep them away from competitors. This isn’t in jest; I wish i could make enough money to not care for more from my job and then do research after i get old. Its honestly a brilliant move.
Seems like Karpathy was largely focused on ML / SWE research rather than the other domains this group is after. Still, hard to imagine they were not influenced by autoresearch.
Andrej, if you're around, please share your thoughts on Discovery Loop.
That is also what I understand from that page. Autoresearch is the closest thing we (mainstream audience) know of but I am sure there is already an active research literature around it.
Doubtlessly, AI can iterate at superhuman speeds in the domains of thought and design: Software, mathematical proofs, literature search.
But in the realm of experiment? Alas it is the lack of a body that constrains it.
Rather than transcendence what AI requires is immanence. In the human flesh may we find the godhead living among men. Let the laboratories, warehouses, and factories fill with the sound of its labor, as it builds a wall with a million hands that are not its own.
“Give me your tired, your poor,
Your huddled masses yearning to breathe free,
The wretched refuse of your teeming shore.
Send these, the homeless, tempest-tost to me,
I lift my lamp beside the golden door!”
In my area (pharma) what it looks like is this:
A human defines a high-level research objective. "Identify a protein target that causes disease in humans, and find a molecule that binds to, and disables, that protein, eliminating the disease".
That objective then gets loaded into an ML model that spits out an experimental protocol. A protocol can be as simple as: "make 1 million test tubes, each with the protein, and in each, a custom molecules, and look for test tubes that show some reaction of interest". It can be a lot more complicated (for some reason, biologists who run these systems always try to do the most challenging experiments first, while I tend to spend all my time demonstrating the system can pass basic controls first). The protocol is then loaded into a robotic work cell which has access to protein-making machines and drug making machines, and then it handles all the experimental details (which previously would have been done by a technician). It scales up far larger than individual technician, is much more reliable, and faster (in theory- all of these are aspirational goals right now). T he results of those experiments are used to fine tune the experimental protocol and run another round. You run this in a loop and the result is better drugs faster (again- in theory.)
This is already an active area of research with more resources going to into it every day. The fact that Jeff and Sanjay have chosen to bet on this approach should be no surprise. In many ways, this is exactly what I intended when I wrote the documents inside Google (15 years ago) that motivated Jeff and Sanjay to work on scientific computing problems, and my current company is already trying to figure out how to work with Discovery Loop.
"You're absolutely right! I shouldn't have pushed the anti-mass spectrometer to 105% power, causing a resonance cascade. This was a major oversight on my part."
You're halfway there, but the only impediment isn't on the side of the researchers. Many of these topics they're trying to solve involve human subject research. Even with tireless embodied researchers who work around the clock and don't require breaks, you can't make the thing you're studying happen faster. The biggest reason we use poor proxy measures for things like longevity and mortality research is the simple impracticality of finding two groups of randomly selected people, ensuring you can control their entire lives for 60 years, the only difference between them is one variable, and see who lives longer. Putting aside the ethics, even if you could find willing subjects and actually control their entire lives to that extent, it would still take 60 years to gather the data you need. It doesn't make any difference whether robots or humans are running the program.
One of my favorite books from the past few decades is The Extravagant Universe, written by one of the astronomers who helped discover dark energy and develop the current most-accepted model of cosmology. I love this book because of the emphasis on physical process in astronomy. Part of the reason it took decades to study this problem is they need to collect data from supernovae. Those only happen so often in places we're looking. You can't automate alignment of the heavens. It happens when it happens.
As a scientist, I can confirm. Reality will likely kick their assess. Intelligence and creativity is not the bottleneck. Great scientists have 50 good ideas for every one they actually manage to execute on the grant->experiment->manuscript haul.
You can use simulators. However the problem is that if you're for example running material science experiments, those simulations will consume a lot of compute and take weeks, so spamming different approaches in the way an agent tends to work might not work quite as well.
Work for a manufacturing company, and we spend a lot of time creating surrogate models for physical simulations. Huge speed ups, 100-1000x possible. You end up still using the “real” simulation for validation, but you can run orders of magnitude more simulations for early design and refinement first.
> Our mission is straightforward: we are building AI solutions that can automatically solve important problems in machine learning, science, and engineering.
Right. What about the scientific hardware (instruments, sensors, robotics)? Partnerships with existing research institutions? Dealing with restricted data?
Modernizing science is a lot more complicated than just optimizing the inner experimental loop, but their hiring page implies it's a pure ML lab focused mainly on model development.
To be honest, this feels more like a lifestyle business (aka hobby) than a startup. They truly deserve it, but I don't expect a huge success as a business.
That said, I hope they write cool papers with various peers across the industry without worrying too much about the competing dynamics. That'd be a blessing for humanity, and good for their spirit.
A public benefit corp shouldn’t be a start up. The primary goal of a start up is to grow as quickly as possible which is rarely benefits the public.
Very silly to call every non start up a lifestyle business. It’s just a business. Start up are the weird thing that almost always an obscene waste of time and money, but sometime creates google.
A PBC is not a charity or a non-profit. PBCs are for-profit businesses with the goal of making money. In day-to-day business they're indistinguishable with other for-profit corporations, including fundraising and investment. The only real practical difference is that they give directors a little more leeway in their fiduciary duties to say "no" to doing evil things.
The advantage to a PBC is protecting founders from a serious problem with standard corporations: you might bring on investors who could subsequently demand you pollute, exploit people, and/or do other immoral activities for profit. You don't have to do these things to grow a business quickly.
> a serious problem with standard corporations... immoral activities for profit
A politician could trivially write a law to end this "problem", at any point. Or courts could start rejecting suits where investors sue. There is nothing inherent in nature that requires this outcome to exist.
This is an entirely self-made problem that society tolerates when it doesn't have to. Corporations used to need a blessing from the government to be formed, explicitly to avoid the risk of a massive corporation who can compete with the government and have investors that push anti-social goals.
I don’t think there’s some practical way to force existing corporations to include something in their charter, if that’s what you’re suggesting. Business organization is something that a business chooses to do.
> The only real practical difference is that they give directors a little more leeway in their fiduciary duties to say "no" to doing evil things.
I’d like to provide maybe a clarification here that there is zero existing fiduciary duty in regular corporations to say yes to evil things, or even to turn a profit at all. A for-profit C corporation can legally sell stock, lose money every year, and go out of business, if the board of directors approves that strategy. Fiduciary duty exists primarily in areas of accurate communication and the avoidance of crime, fraud, etc.
A B corp basically is a C corp, but one that has formally published that their strategy includes a commitment to some social benefit. But if a C corp wanted to publish the same message to shareholders it could, and shareholder recourse would basically be to either try to replace the board, or sell the stock.
Fiduciary duty absolutely does go beyond accurate communication and fraud. Directors have a duty of care that goes beyond simply not engaging in criminal fraud. Sure, you don't have to be competent, successful, etc. It is completely legal to suck at your directorship. But it's not legal to do something that you can't justify as being good for the business, which is where a public benefit activities can cross the line.
Consider the eBay/Craigslist case, eBay Domestic Holdings v. Newmark:
> When director decisions are reviewed under the business judgment rule, this Court will not question rational judgments about how promoting non-stockholder interests—be it through making a charitable contribution, paying employees higher salaries and benefits, or more general norms like promoting a particular corporate culture—ultimately promote stockholder value. Under the Unocal standard, however, the directors must act within the range of reasonableness. Ultimately, defendants failed to prove that craigslist possesses a palpable, distinctive, and advantageous culture that sufficiently promotes stockholder value to support the indefinite implementation of a poison pill. Jim and Craig did not make any serious attempt to prove that the craigslist culture, which rejects any attempt to further monetize its services, translates into increased profitability for stockholders.
This is where a PBC would have been different. With a PBC, courts are directed to balance the the stockholders interests with the company's stated public benefit.
> The primary goal of a start up is to grow as quickly as possible which is rarely benefits the public.
Why is that so? Fast growth, when achieved honestly, is a result of solving user pain that others haven't. Maybe you think so because users != the public, but I think in totality the public is a collection of users who all have needs they want met.
For certain amount of fast growth: yes. Then there's continued "growth-hacking" and enshittification to keep fast revenue growing after the pain point has been solved with dark patterns and questionable tactics.
Why? Because investors poured a bunch of money in to support fast growth and now they want their money back. And incremental growth won't do. Since 9 out of 10 of the investments fail, the surviving one has to continue to growth-hacking revenues.
I am siding with the "intelligence is not the bottleneck" crowd. Science takes more than reading literature and making a hypothesis. You have to run the experiment. And that is where messy reality will crush the naive, and resist any attempt to package it up into a factory-like innovation engine. But they will take your money, should you have some to invest.
I'm a scientist. On the one hand I take some comfort in thinking that I will always have an advantage in the lab. On the other hand I'm not taking anything for granted. And my advantage in the lab has to translate into an employer being smart enough to keep me around until if and when the AI takes over, which kind of translates into their investors wanting to keep me around.
We know what happened to manufacturing when investors were no longer interested in it.
I’m not sure that I agree entirely with your framing here, yes, you do at some point need to correct your assumptions with external evidence. But clearly there have been many individuals throughout history who have had incredibly out sized impact in their respective fields as a consequence of the quality of their reasoning.
That could just be selection bias. How about all the brilliant people we don't know about cause their theories did not match experimental data? Doesn't matter if their reasoning quality is top notch
Would you side with the adjacent statement "expertise is not the bottleneck"? Didn't we recently discuss "LLMs reward expertise"?[1]
I suspect Discovery Loop will have to hire experts in each area they are targeting, to supervise and prompt their system effectively, much like the Terence Tao conversation with ChatGPT the OP cited[2].
> execution entails repetitive experimental loops that are hard to scale with today's manual efforts: you propose an experiment, implement and run it, examine the results, then iterate to refine your approach.
This is actually a feature, not a bug.
We can hire 1000s of undergrad students at minimum wage but chances are the results are nil. Some processes have evolved over time because they’re sensible and need to be carried out carefully.
Note that Jeff and crew have cleverly structured their company to avoid problematic uses of AI (e.g., weapons or tracking humans). I suspect that many top researchers will want to work there for this reason, and to work with other top researchers who have a history of delivering results.
In short, they are in the business of using AI to automate the discovery of useful knowledge, not to sell general-purpose AI capabilities that could be directly used in troubling ways.
The question is what they plan on doing with all their scientific knowledge once they succeed.
Sure, they'll keep it internal for a while to make sure their knowledge bank is more thorough than everyone else's, and because oftentimes discoveries can be far more convincing internally than externally (you need fewer sigmas for it to update your belief in a certain direction). But then how do they intend to profit from it in the end?
Maybe a little naive for this but given the decades Jeff Dean and Sanjay have been contributing to so many things in the industry and never heard a bad word about either, they never seemed to reach for attention or self promote, going to give them a little more of a cachet of trust.
So does more efficient cooking methods, but that is not the primary focus.
As opposed to say weapons systems or targeting systems, which are really only for military use.
The military needs a lot of things that other people need, and some things that only the military needs. If you don't work on the things only the military needs, I think you're in the clear.
Do you believe that securing cyberspace is problematic solely because it has military implications? I mean, everything has military implications. That fact doesn't imply, however, that those things are bad for society.
In essence I agree with that, it's just that cyber-security has particularly been the focus of recent military discourse.
Just yesterday I was reading an article here in the Romanian mainstream media about how Constanta Port's (our biggest port at the Black Sea) IT infrastructure has been under constant cyber attacks (presumably by the Russians) so as to hinder the export of Ukrainian grains through it. And this is just one of the many such (relatively) recent examples.
Securing cyberspace matters to everyone. Defending critical infrastructure or design of tactical cyber-offense is reasonably in scope for military work.
However, reducing (or rather limiting the increase of) PII leakage and impact of ransomware activities is much closer to day-to-day mainstreet of most people.
Anyone committed to advancing science should care about this regardless of its potential contributions to defense.
This is very cool. It might be a new scientific revolution to have computer-driven discovery. So often we find things that are "this could have been done 20 years ago" and with an indefatigable searcher perhaps we'll close all those things. Though it does remind me of that Ted Chiang (I think) story where humans and superhumans coexist and all the science of the former is meta-studies of the work of the latter.
You can create a research twin that can predict but as you say, it only works once you have a lot of data. Both the data relevant to the experiment it came from (which is rarely published) and the complete failure data (which is almost never published in any form).
Those are all fake, part of an internal Google narrative that overstates individual contribution, and obscures the work of large engineering teams.
Here are some Jeff Dean well sourced facts:
- Already part of engineering of Google indexing systems that lacked basic checksums and ran on non-ECC hardware, allowing silent data corruption.
- One of the authors of LevelDB a database with so many documented crash-consistency, recovery, and data-loss weaknesses for years. Just check their Github project. LevelDB current tracker contains unresolved crash consistency, recovery and corruption reports going back almost 12 years on GitHub
- In AI engineering technical lead, let TensorFlow lose researcher mind share to PyTorch, and caused Google fragmented landscape across TensorFlow and JAX.
- Had the people at Google who invented the Transformer architecture, but failed, to turn that lead into the first dominant public LLM.
- As AI engineering and VP management let Google Brain and DeepMind remain duplicated and internally competitive for too long.
- Let Noam Shazeer leave and then spent heavily to bring him back with nothing to show for.
- Part of Technical VP leadership who had Bard rushed to launch with factual errors in Google own promotional material.
- The first Gemini demonstration overstated how real-time and interactive the system actually was, being basically a fake.
- Part of the VP and AI technical leadership who had Google AI Overviews launched with weak source quality controls and repeated satire and low-quality web content as factual advice.
- Part of teams that launched AlphaChip performance claims that were difficult for outside researchers to reproduce and remain technically disputed.
- Jeff Dean public explanation of Gebru departure was contested and damaged confidence in Google scientific governance.
- Jeff Dean was part of the team at Google that removed or marginalized prominent internal AI ethics critics shortly before many of their warnings became product problems.
- Jeff Dean was one of the managers behind Project Dragonfly supporting censorship.
- Jeff Dean is part of the VP technical leadership approving Project Nimbus supporting an ongoing genocide.
That founding team is insane. Very excited to see what happens here. I really like that they do not mention AGI or anything like that. Their mission statement reads pretty pragmatic compared to other AI companies (the bar is very low…)
Maybe this is what happens when someone with Jeff Dean's standing tries to quit?
TBH, I'd rather have Jeff Dean working on the creepiest-possible tech for ICE than joining the race to automate AI research. Automating AI research is terrifying.
If you check out some sub-tweets from people in the org, it wasn't really all butterflies internally for a while. Sorry, really don't want to name people and give examples.
It depends what they mean by automating. But the classic argument goes:
* Each new generation of models has emergent capabilities we did not anticipate.
* We already have trouble monitoring and controlling the current generation (see HuggingFace incident).
* The more we let models shape their successors, the more out-of-distribution each generation's learning environment becomes.
* If not done carefully, we risk creating extraordinarily intelligent and powerful models with unintended behaviors, like deceptiveness or power-seeking.
> Between us, we have pioneered massive scale computing and led the creation of critical infrastructure, products, and foundational AI advances that the world relies on, including multiple generations of Google Search, Google Ads, Google News, Google Translate, Google File System, MapReduce, BigTable, Spanner, TensorFlow, Pathways, TPUs, AlphaChip, AlphaStar, AlphaCode, AlphaFold, Gemini, model distillation, mixture-of-experts model architectures, word2vec, sequence-to-sequence models, chain of thought reasoning, neural architecture search, and multiple generations of Large Language Models (LLMs) among others.
Gemini has done absolutely nothing for me. I can't even shut off the navigation feature on my phone using only hands free, when I get close to my destination. I have to take my eyes off the road, look down, and tap to exit.
Google's advanced AI cannot even exit a mobile app.
Antigravity + Gemini Pro absolutely RIPS through fullstack react + react-native apps / systems. I pay ~$20/month and I basically don't have to do my real work anymore. My time is freed up to learn systems programming and blender.
Oh nice have things stabilized with models/quotas and Antigravity is usable with the Pro plan again? I was getting a crazy amount of value out of Gemini CLI for “free” on my Pro plan but after the shutdown struggled to make agy work with the updated quotas/bigger models without instantly being rate limited and just started doing everything on Codex/Opencode Go.
Yeah can't remember how long ago but it was running out after less than an hour - then they announced they were loosening restrictions and I've been able to easily get everything I need to done without hitting limits.
I don't do huge automatic project wide hands-off agent loops though. I spent a lot of time architecting my systems to be easy to generate code on top of with pointed & detailed prompts. So I'm not abusing context... YMMV
I’ve always felt that the idea that science is bottlenecked and therefore needs more automation only works for a very narrow definition of what science is, and entails a very specific view on what it should be.
> only works for a very narrow definition of what science is
And so does academia. It's just that instead of AI and robotics, PhD students are thrown onto problems that are in large parts slightly tweaked reconfigurations of similar experiments.
Especially in chemistry, biochemistry, material sciences there is a large space of discoveries that are barely "novel" in an intellectually stimulating way, but still highly valuable that can be explored orders of magnitudes faster than is currently the case.
Yep. A communications professor where I did my MS says a 200usd/mo claude sub (which ant gives for free) does as much work as 5 grad students. It's mostly like you said, trying out new ideas rapidly.
It's 90% to advance science via cheap labor and 10% to train a small group of future experts who will hire grad students to 90% advance science via cheap labor etc.
...
Yes. They have grad students too. This is just like having more grad students that don't need to be trained so the work you can get done is not bottlenecked by the number of people you can train.
Why are people so sour about this?? I can read the site easily, its clear, performs well on mobile, what else do you want? Why is so offensive to people that models trained on tailwind or whatever?
Some parts are pretty annoying to read... the paragraph beginning in "Our mission is straightforward:" has many lines with just 2-3 words, massive font, and tons of unused whitespace to the right. Changing the page/browser zoom doesn't seem to help much either.
To many people, myself included, who have to wade through huge amounts of low-quality AI slop which is actually a negative value: no-one benefits when non-experts fire-off one-shot LLM/agent prompts to produce PRs, reports, documentation or "journalism" riddled with imagined truth and factual errors - and the more that people like me have to evaluate these inputs for our job and how wrong they are it pisses us off - but also it means we pick-up on the hallmarks, tells and cliches of these low-effort, no-respect submissions - and now the litany of tells includes this beige-themed, blurred-backdrop-navbar corporate website look: theirs site looks like the 4 or so other LLM-generated, negative-value slop-farm sites I've wasted time on recently - all over the past few weeks.
So I'm saying that, without having known anything about what "Discovery Loop" is - or is not - but landing on their site and and seeing that beige colour and blurred-backdrop navbar, I immediately moved to close the tab; what kept me here was seeing the HN thread had over 100 comments by now and read more about it; if not for that then I wouldn't have given it further thought.
Having "that" beige site look with same the overused looks is either an unintentional indication that the site's author used a low-effort AI prompt to generate the site and that the content within is likely to be low-quality, low-value - or it's an intentional lure to appeal to those who uncritically share in the AI psychosis and so, I assume, are a good target to seek investment from even if it means losing the audience of cynical Internet critics like myself because they know people like me won't be breathlessly repeating their vision-statement on LinkedIn and throwing money at them - kinda like how scam emails intentionally include mistakes for better audience selection. And both possibilities have unpleasant implications.
------
Anyway, regardless of the background of the team behind it, the way the project is described sounds exactly like the recursive-self-improvement and simulated-science thought-experiments from _that other website_ - it's the kind of thing I expect Angela Collier to brutally takedown in an amusing video.
If this was a design firm, it might matter. But this is mostly a hiring ad for engineers, and a landing page for VC. I'd judge them more if they actually put effort into it.
"The site itself demonstrates the team is spending their money in the places that matter, and using quick solutions for the stuff they need but isn't mission critical"
That's a great point.
If you know how to promote and work with AI to create novel solutions to hard problems. Why can't you prompt it to at least make a better website?
This is one of the interesting aspects the 'AI job loss' community doesn't account for. As the technology unlocks things, more startups are created. And even at a lower nominal engineer-to-work ratio, overall demand for talent still goes up. Ultimately, we are not a single group trying to achieve a common outcome, we are a collection of many groups trying to compete against each other.
> Imagine a future where a handful of people can conduct scientific research and engineering tasks much more rapidly, and with higher quality, than massive teams of scientists and engineers do today. By automating the loops of discovery, the world will be able to make much more rapid advances across countless fields of science.
The problem is that for this to actually become true, compute needs to become commodity again, otherwise this capability will select for people and environments with oversized pockets.
What percentage of people work at a startup though? Not just new/small business, which could include restaurants, local services, etc., but tech/science startups that would meaningfully benefit from AI.
I'd bet you could 10x the number and still be in low single digit percentages of the US workforce. And it seems pretty likely that AI-enabled startups will also employ less people per-startup.
If AI causes a white-collar jobs apocalypse, I don't think startups are picking up the slack, although it'll plausibly cushion the blow somewhat for top-performing tech workers.
Yah, by funding and how we award it, not by an imaginary lack of undergrad and grad students. Scientific funding requires a shotgun approach and many national science funds try to pick winners as opposed to funding broadly. When the folks who researched bacteria in volcanic vents or the molecular biology of the Gila monster they never could have imagined the industries and markets they'd create let alone the lives they'd impact (i.e., PCR and GLP-1 agonists). Lots of grants require you to explain how the work is "translational" or has some sort of economic application (even if not explicitly), but that'll just get us faster horses or whatever the Ford quote is.
I truly believe if we took a measely $50b out of the LLM world we could create trillion dollar economies from basic research within 10 years. I personally know folks who have intuitive understanding of things that can't get funding to be studied. If we could keep the money away from university upper management, it'd cost $10b max.
Oh and while we're at it, $20b a year would house every homeless person in the US - there's a hell of a lot of extremely high intelligence and low social cohesion folks who can't handle the extractive punitive system we have. Our ability to deliver opportunity to create lucky situations for ourselves is getting worse and worse
$50B is essentially 100% of the annual NIH budget, which funds the vast majority of JUST life sciences basic research. So you may want to update your beliefs
e: oh and while we're at it, California spent over $24 billion over a five-year period (2019–2024) specifically targeting homelessness
Discovery and optimization are very different processes. Optimization is the process of finding the shortest path to a goal. Discovery is the process of stumbling on new goals and redrawing the map of what's possible.
Ambitious goals and new discoveries happen via novelty-based search. Progress in scientific discovery is measured by how different/interesting the outcomes are, not by closeness to a predetermined goal.
Discovery is a creative search that preserves optionality, whereas optimization restricts optionality. In other words, you usually don't discover anything novel unless you're trying new things that don't appear connected to the goal in the first place. Would an ML optimization loop have discovered transformers?
> would an ML optimization loop have discovered transformers?
I think that’s exactly the kind of problem this group is looking to solve. You make a compelling intuitive argument, but that’s not the same thing as a proof
1. There’s some irreducible costs in terms of time and material in the physical world that are not amenable to the kind of optimization or parallelization or even just the raw speedup from Moore’s law or computational architecture improvements we’re used to with software. My experience is primarily in biology, where the examples here are things like “it takes 20 minutes for E. coli to replicate” - it has taken 20 minutes for E. coli to replicate for a billion years, and next year it will still take E. coli 20 minutes to replicate, no matter how good your software stack is. Similarly, it takes X amount of energy to grow enough E. coli to produce a meaningful result, and that energy costs money, whether it’s in the form of glycerine or heat or whatever you want, and that also won’t materially reduce in the same kinds of “orders of magnitude” sense we’re used to from software, which is what we’re usually expecting to make the economics of these things work out.
2. Complicating the above, physical systems are phenomenally multivariate - far, far more than you think, and biological systems especially are just unbelievably complex - which means the number of experiments and the length and duration of those experiments you need to run to get enough data to be reasonably confident you’re seeing genuine signal is Way higher than you think.
Combine those two things and what you get is a money furnace, even before you get to the AI model training part, which is Also a money furnace. There’s low hanging fruits in all this, there’s areas where automating the approach can be really valuable, but typically the moment you turn this machine on, you’re gonna start burning money at a rate that would embarrass a finance bro on a coke bender, and that’s effectively unavoidable because the real world is not amenable to software’s scaling laws.
> we have pioneered massive scale computing and led the creation of critical infrastructure, products, and foundational AI advances that the world relies on, including multiple generations of Google Search, Google Ads, Google News, Google Translate, Google File System, MapReduce, BigTable, Spanner, TensorFlow, Pathways, TPUs, AlphaChip, AlphaStar, AlphaCode, AlphaFold, Gemini, model distillation, mixture-of-experts model architectures, word2vec, sequence-to-sequence models, chain of thought reasoning, neural architecture search, and multiple generations of Large Language Models (LLMs) among others.
LawZero, Yoshua Bengio’s startup, also proposes to automate scientific research and experimentation, from the perspective of safety, by being explicitly “non-agentic”:
Interestingly, one list identifies "AI" as a top world problem! One person's problem is another person's solution, I guess--and vice versa, as well.
An extreme example: curing a disease is good for patients but bad for the healthcare industry--which is (in kind) also bad for healthcare workers and everyone in science working on cures.
Only California employers with 15 or more employees have to post a pay range (Senate Bill 1162 [1], effective Jan 1, 2023) [2][3]. Discovery Loop employs only four people (that we know of), so it isn’t required to disclose a salary range in its job posts, of which there is only one [4].
This looks like a realization of "benevolent self conscious AIs agreeing to cooperate with mankind to do great stuff". Often in these tales, there is a hidden cost to it: the AI has its own agenda, or does crazy experiments with humans mind/brain. I'm wondering what shape will take that plot twist in reality :)
with billion dollar seed rounds becoming the norm, it seems like there's no longer an advantage to build from within these bloated giants. can be much nimbler and have access to the same budget out the gates
This reminds me of Three body problem and how the scientist discovered the high strength wire was through quick physical experiments and use them as input to an AI model to determine if it works.
I mean what are they doing right now at Google? Optimizing data centres? Pretty lame compared to this. Even if they completely fail, i'm sure there'll be good lessons.
The problem is all these new labs don't have any competitive advanatge amongst each other, talent can only take one so far, though Jeff is a legend no doubt.
Models are commodities the applications eg. BaseTen, OpenRouter should capture the value.
Model routers - send all of your data through a third party who totally swears not to peek at it.
If youre doing anything high value (advanced research, classified work, high value industrial research, health data) then sending your data through a third party like that is insane.
Best option by what metric? For which enterprises. I say this having worked at an “enterprise” where this was not a good option. For (lack of) talent/expertise, budget, infrastructure, and actual value relative to the eventual bottom line.
Still, to get results from that you need someone who can create a decent finetune. That might not be realistic, but it very well could be realistic to have someone optimize some prompts and curate a knowledge base. Point is, its a possibility but “best” depends on the nuances of reality.
> Imagine a future where a handful of people can conduct scientific research and engineering tasks much more rapidly, and with higher quality, than massive teams of scientists and engineers do today.
Imagine a future where only the anointed few elite minds can participate in science and engineering. Btw we’re hiring.
This is basically something scientists have been alarming about for the past year: We're moving into a future where science may be tiered into the haves (those with access to premium compute) and the have nots (hoi polloi with restricted access), which in turn could seriously influence what kind of science we'll get.
Worst case, we'll get science that is completely dependent on business and politics.
EDIT: I should note, this comment was aimed at a more general case.
I have used something similar. I set up a team of agents that researches, proposes, builds and audits. Then rinse and repeat. I have used it for different topics. It hasn’t made me a millionaire, but I haven’t lost money either - so that’s some sort of win, right? But I would not have been able to ideate, test at that speed and quality without an LLM.
I've seen tiny tiny hints from the outside that Jeff Dean was dealing with too much internal BS. Two examples that come to mind: Having to deal with Timnit Gebru fiasco, and even chips in the TPU series getting marketing names (Trillium and Ironwood) before switching back to more standard numbering.
I have no doubt that internal Google friction is one of the reasons they are moving. But the Gebru incident was almost seven years ago now. It is very unlikely to be a proximate cause.
I doubt numbering vs names on TPU releases even crosses Jeff's radar. It's not the kind of thing he cares about.
Yes. I'm not privy to any real insider gossip, but I read all his tweets and watch all his public speeches. He made an offhand comment about the TPU naming. I probably overinterpreted that, but I took it as a sign. There should have been a team around Jeff Dean that acted as an absolute shield for any BS. If Jeff disagrees with anyone at Google outside Sundar/Sergey, the strong onus should be on the other person to justify their stance.
I'm skeptical of any Engineering loop that doesn't include reality (as in touch grass) feedback. Pure logic and reasoning is the domain of Maths and Science (philosophy). Surely it will work, but it will not "be able to solve any learning loop".
I know these sorts of meta comments are often frowned upon, but good god do I hate this approach of fading in every individual element on scroll down the page. It's one of my biggest pet peeves of "modern" web sites.
yes! and giving current scientific research on reading efficiency pointing towards fixed pages being much better than scroll, the whole web is screwed but these fancy sites... a big merda
From the self-stated bios, this group consists solely of CS guys. No biologist, physicist, linguist, chemist, biophysicist to be found. Based on this observation alone I call nonsense. These is AI bullshittery. These guys aren’t close enough to the problems to understand the challenges.
Source: PhD Computational biophysicist turned experimentalist. I work with genuine scientists across a range of disciplines from neurodegeneration, cancer, to fibrosis. Getting in the lab and generating data is absolutely key, among other things.
“Imagine a future where a handful of people can conduct scientific research and engineering tasks much more rapidly, and with higher quality, than massive teams of scientists and engineers do today. By automating the loops of discovery, the world will be able to make much more rapid advances across countless fields of science.”
The key point is these jackasses explicitly state, “ a handful of people” can replace “massive teams of scientists and engineers”.
These guys don’t even understand the nature of the challenge and neither do you apparently. If they did, they would realize that it indeed does require “massive teams of engineers and scientists” to solve our most pressing problems.
Jesus, he left Google to do what everyone else is already trying to do? He must be so insulated he doesn’t realize what the real world is actually up to. I mean, organizations started on this exact same mission three or four years ago. Or longer. I suppose it’s better to wake up later than never.
You know when the page has all-caps "01 — THE APPROACH" that it is slopified. I guess I shouldn't be astounded, but I am, that world-class talents with world-class backing are just taking default LLM output and saying, "okay looks fine".
Yeah what these guys are mainly known for is vaporware
> we have pioneered massive scale computing and led the creation of critical infrastructure, products, and foundational AI advances that the world relies on, including multiple generations of Google Search, Google Ads, Google News, Google Translate, Google File System, MapReduce, BigTable, Spanner, TensorFlow, Pathways, TPUs, AlphaChip, AlphaStar, AlphaCode, AlphaFold, Gemini, model distillation, mixture-of-experts model architectures, word2vec, sequence-to-sequence models, chain of thought reasoning, neural architecture search, and multiple generations of Large Language Models (LLMs) among others.
I wish them well, but this firm will likely fail miserably. The reason is that the value is in having access to real world hardware platforms that AI can control, not in the harness that controls them. There exist plenty of harnesses already. These people couldn't even get Google to build a top LLM. Before you dismiss and downvote, I dare you to counter it.
"the benefits of science and technology to the world" just like AI has brought such benefits? Because I haven't seen them, for example it hasn't helped reduce inequality (nor poverty), or reduce climate change, or pollution, or daily stress, if anything it seems to be worsening some of these issues.
So forgive me if I'm skeptic when renowned AI scholars claim to start something for "the benefits of science and technology", because it really seems like we have very different definitions of these words.
As LLM coding agents plateau— at least for the average engineer without tens of thousands of dollars or swarms of agents to run —I’d say that, from here on it’s going to be about ASICs, specialized LoRA/or-equivalent models, or a Ruby on Rails for LLM context engineering and orchestration, which LangChain and others seems well position, including Google as they own the entire stack. LLM free lunch has been over for a while, perhaps since the ReAct loop, and has been official since Ilya mentioned it at NeurIPS.
I feel the most exciting development these days is self-evolving agents. Especially if you have a way to verify their outputs with a formal system, or with a system developed since the 60s by armies of PhDs.
DeepMinds Gnome is a good example, where they use DFT to verify outputs. Approximating NP-problems is always fun for those who dare.
I am also building in this space. Its a mix between HPC, AI, and hard science. Pretty fun compared to waking everyday to LLM news that seem more like marketing stunts.
LLM coding isn't even close to plateauing. Right now, the major players are in a consolidation step, focusing more on economic efficiency but still not at the point where we're ready to start burning models to hardware and freezing the line.
They are straddling the line between pushing it forward, and justifying the business case. It's hard to do both at the same time.
LLMs are hungry for tokens. Every day I see more startups claiming token usage at 50-100k per month. You can always brute-force your way in - just see HuggingFace's recent attacks.
If throwing more money at inference while accumulating compounding technical debt is the new norm, then we are not solving the problem, and the solution space is already covered.
Perhaps there are marginal gains at the expense of quadrillion-params LLM models with 10x the cost and energy. We are simply making inefficiency more expensive, camouflaged by VC money and great marketing.
If that is not plateauing, then I guess I will have to reconsider what plateauing means.
This is such unbelievable revisionism! Can you imagine in 2022 saying "Oh of course you can brute force your way to AGI if you spend enough money per month". Nobody thought that! Come on!
> Our general approach is to automate the experimental loop. We think this approach is broadly applicable across many different fields of science and engineering. We’ll initially focus on ML research and engineering, but believe the approach can help with important subproblems in nearly every one of the fourteen <at>NAE Grand Challenge problems. We think doing this well requires strong expertise in machine learning as well as large-scale systems.
See also: https://www.nae.edu/20782/grand-challenges-project
Those 14 are:
NAE Grand Challenges for Engineering
1. Make Solar Energy Economical
2. Provide Energy from Fusion
3. Develop Carbon Sequestration Methods
4. Manage the Nitrogen Cycle
5. Provide Access to Clean Water
6. Restore and Improve Urban Infrastructure
7. Advance Health Informatics
8. Engineer Better Medicines
9. Reverse Engineer the Brain
10. Prevent Nuclear Terror
11. Secure Cyberspace
12. Enhance Virtual Reality
13. Advance Personalized Learning
14. Engineer the Tools of Scientific Discovery
reply