Former Google CEO Eric Schmidt's Leaked Stanford Talk

i believe this trend is a result of management culture and performance metrics that attempt to measure "impact", and correlate pay and promotions along those measures.

It's the same reason why google products die when they don't reach mega-success (thus products getting killed off in spades recently).

Nobody wants to be doing maintenance on projects started by somebody else - so as soon as the lead visionary leaves for better pastures, that project gets languished, and whoever takes over it cannot use it to generate promotion worthy impact.

> tensorflow looks like currently loosing to pytorch - seems like google got bored and more development is for JAX, Keras wrapper

Well, TensorFlow doesn't "look like currently losing", it has already lost since a long time. I haven't seen a decent paper release code in TensorFlow in years, and all the references I see to TF online are job posts from "older" companies (to the point that, if you are looking for a job in data science, seeing TF mentioned in the job post is kind of a red flag of a place you don't want to be).

That said, I am quite certain that this has only a small impact on why Google is losing terrain, and even on why it is behind in AI (which is also debatable: narrative aside, Gemini is not that much lacking behind competitors). Certainly if TensorFlow + TPUs turned out to be better than PyTorch + GPUs they would have had a lead to start from, but if that was so important, Meta or NVIDIA would have created the first LLM, not OpenAI.

Simply, sometimes stuff happens, you can't predict it all.

This is a weird take. It might pertain to 2020 during Covid when Meta encouraged remote working, but it's not like that anymore. Interviews were always leetcode focused so nothing changed there.

Meta is an interesting comparison because the performance culture has historically been way tougher than Google (sans Covid years). It's quite a shock for a lot of Xooglers who join. One Xoogler at Meta who told me they only used to work 4-6 hours a day at Google. That is basically impossible in Meta - you have to constantly prove your value and the perf bar is high.

I think Google is changing now though and I wouldn't be surprised at all to see them adopt Meta/Amazon style perf cultures in the future.

Source: I work at Meta, and have close friends at Google.

Tensorflow is a symbolic framework, which is less intuitive to work with for most people than the Pytorch. Not to mention the errors Tensorflow generates are more annoying to debug (again more an issue with the fact that it's symbolic than any lack of effort on part of Google)

Google tried to fix it by introducing an eager mode in Tensorflow but by then it was too late.

Typescript is a superset of javascript, Dart isn't (though it transpiles). Flutter implements its own widgets, react native controls native widgets.

Google seems to have a deep-seated distrust for programming language theory, as well as a deep-seated distrust for its users. This combination produces awkward software and APIs that ignore modern PLT and take a "Google knows best" approach.

Yes, that's why you go to large megacorps. I wouldn't go to Oracle, IBM, Google, Facebook, MS, etc. etc. if I had a "huge fire or a drive to win". Honestly, I go there because I wanna work 5 hours a day (less if possible) and have a stable career.

If I want to work 9-12h a day, give me an upside. None of these huge megacorps will reward that.

If I'm the lead of a Goog project that becomes a hit, do I get $10M bonus? Of course not. I get a pat on the back and something to put in my packet for the next promotion interview.

So you're absolutely right, but the problem is not in people. It's in the way the system's designed.

Google has severe cultural problems and the responsibility for that should be assumed by the leadership.

I don't agree, but there are challenges at Google that don't exist in other large companies.

Google's monorepo along with the 'bottom up' driven culture means that to enact any non trivial change, you need many sign-offs, which means an iron clad design doc, and either you have to be brave enough to make huge disparate changes across other teams codebasse, OR to wait multiple quarters to have respective teams execute.

Sometimes this is awesome, and it means you can do really complex things over the course of a year.

Often, it just means experimenting with non trivial changes takes 3x longer than it would at say Amazon or more silo'd companies.

No one with such a drive would stay in, or even go to Google, neither any GAFAM-like corp. Those have been the establishment for 10+ years now.

A large corporation cannot seriously pretend to encourage their employees to "play to win" when it is structuring itself against that precisely.

The bigger the corp, the slower, duller and more rigid it becomes _unless_ someone at the CxO level actively fights against it in a opinionated manner (which still doesn't often register well with boards and major shareholders).

It may be sad, it's quasi a law of physics for corporations. If you have that kind of fire in you, go first hunt into to the woods, don't go to the factories.

People love to talk about how capitalism (or free market) solves the tragedy of the commons, but it actually doesn't. The only way to solve it is to make people care about commons by giving them a say in managing it, i.e. socialism.

https://qz.com/work/1326942/sergey-brin-started-google-with-...

Yep, he is a great leader! :) /s

[1] https://www.nytimes.com/2024/07/12/technology/eric-schmidt-a...

There's no incentive to excel. Either you'll get laid off anyway or someone else will take credit for your work.

Might as well just coast. The severance is quite good, so there's not even much motivation to quit, better to wait to be laid off.

I think Google is starting to become that. It's just the nature of older very large corporations, I suppose. There are places to hide in these companies for non-productive workers.

I'm not trying to defend Schmidt, but you're naive if you think the US shouldn't develop AI weapons.

How would the world look like now if the US took the high road and refused to develop nuclear weapons and instead let Nazi Germany and Soviet Russia develop them instead?

no, you just need to vaguely know what you want, and get the LLM to produce something that you then examine, and crawl towards the end goal.

LLM's could potentially allow fast iteration from a laymen's description of what they want.

The other day I was looking into adding Trusted Types in the Content-Security-Policy header, which was something new to me. In my chat with Claude I asked:

"Lets brainstorm 10 a list of ideas closely related to this so we can think of anything we might be missing on the topic to consider."

And that provided a good list of items to review to consider and expand out the sphere of thinking for the LLM.

It is an infuriatingly hard problem to have the LLM produce excellent results every single time, and have it just do everything and want it to read our mind and all the knowledge and context of a task. I think we'll make some good progress over the next few years as agentic workflows are built out to mimic out thought processes, and the cost/capability of the LLMs keeps improving.

Something to keep in mind when reading his other pronouncements.

The use of AI/robots in war is probably not a good way to ingrain into people how to avoid conflicts either: as sad as it sound, let them get a taste of it first hand, and that'll probably calm things down more efficiently: not for a few hours/days, but decades.

> This is not actually one of the things to know about war. It's about as far from the truth as you can get.

... Unless you are writing the rules of war:

"Secretary of Defense Ash Carter appointed Schmidt as chairman of the DoD Innovation Advisory Board announced March 2, 2016. It will be modeled like the Defense Business Board and will facilitate the Pentagon at becoming more innovative and adaptive.

In August 2020, Schmidt launched the podcast Reimagine with Eric Schmidt.[71][72] In December 2021, Schmidt joined Chainlink Labs as a strategic advisor.[73] In October 2022, he co-authored a piece titled "America Could Lose the Tech Contest With China" for Foreign Affairs with Ylli Bajraktari, former executive director of the U.S. National Security Commission on Artificial Intelligence.[74] In March 2023, Schmidt testified at a U.S. House Committee on Oversight and Accountability hearing regarding AI.

In 2022, Schmidt was appointed to the National Security Commission on Emerging Biotechnology, a legislative commission charged with making policy recommendations to Congress and the Executive Branch.[1]"

"Since 2023, Schmidt has been involved in building White Stork, a startup developing suicide attack drones.[2]"

[1] https://en.wikipedia.org/wiki/Eric_Schmidt

[2] https://www.forbes.com/sites/sarahemerson/2024/01/23/eric-sc...

Iran just sent I-don’t-know-how-many drones and missiles at Israel like a few months ago. A few of them landed, but most were caught and intercepted in the air. Here’s the thing: if even one of them had hit, say, the center of Tel Aviv, or the old city in Jerusalem, it would have been a massively successful attack, even if none of the other ones had done any damage. The size of zero armies was measured in that exchange.

Normal warefare is based on "Manpower" and you reference "Army". He further goes on to explain that artilery, tanks and any kind of "land army" is completely obsolete.

In war, defenders can entrench more and more and a lot of work and planning is put into hitting either before the defenses are up, or not at all.

I don't agree with US involvement in Iraq, Afghanistan and many other places, but Ukraine is a obviously totally different from those. Marjorie Taylor Greene may not give a damn as she lives thousands of kilometers from Russia (or maybe she's just been bought by Russian money), but morally she's wrong. When it comes to Russian imperialism, the only right thing to do is help its victims fight back.

> A future where Chinese companies own the EV market in the US with all of the American user data going back to servers in China- that's just not a future the US govt. wants

I agree, but why do cars need to send data to servers again? I don't grok why Tesla does it and I don't grok why BYD does it either. What's wrong with the alternative where we allow real competition that benefits everyone, and then ban any features (eg sending data to servers) that the government gets nervous about? I don't see how this is any different from the myriad of other requirements on cars (eg mandated safety belts etc etc etc).

I recognize that state-subsidised competition still isn't fair, so I support import tariffs that match those subsidies, but any further and you're just hurting competition and the consumer, right?

Thing is, super cheap EVs and solar panels is exactly what we should want. There's a climate crisis going on and these things are a big part of the solution. With climate goggles on, to put import tariffs on the tech that's gonna save the world sound batshit insane. We should want those tariffs to be as low as possible.

This [video](https://youtu.be/BQ23sgi_mgw?si=3HOm3WeWKO7VJMzr) called <Is China’s High-Tech ‘Overproduction’ Killing Jobs In The West?> from Channel NewsAsia covers it. Highly recommend you westerners to watch it; there's some really heavy handed defense/reasoning from the chinese side.

i say china is not at all responsible for the drug, except as a contract manufacturer. If china doesn't do it, somebody else would've. China just does it cheaply.

The drug problems come from the cartels, and from the fact that US's war on drugs makes drugs expensive, and thus profitable. There's no possibility of humans relinquishing drug use - think about the prohibition and the problems that it caused.

I say, just allow drugs to be manufactured, and sold (under license and regulation) in the US. It's gonna be like alcohol. This will take away the profits from the cartels, and they will stop buying from china. It will increase tax revenue, as drug consumption can be taxed (like tobacco today is).

It will also remove the stigma, and reduce drug enforcement costs in police. The only thing society has to give up is the moral high horse of drug use.

> all of the American user data going back to servers in China

There's a long running argument with the EU over all EU user's data going back to the US where there are no privacy laws or protection against surveillance (US constitution only protects US nationals!), so the US government would have more credibility on this if it recognized that.

How about a future where user data goes to no servers of no country at all? Am I asking too much?

How many independent innovations get snatched (sherlocked?) by big business taking the idea and making their own version, when they see a small player that found a customer segment a large company never bothered investigating?

When large sums of money is what’s needed to execute any novel idea to full completion before your competitors can, that money becomes just a version of ”IP theft” or unfair competition. I think what they are getting is a taste of their own medicine.

EVs were a mature idea a decade+ ago, but big money oil companies hamstrung it and big money ICE manufacturers argued it was not worth their time or effort. The Chinese market and leaders knew the fundamentals were valid and cornered the entire manufacturing market before it could be cornered by western companies.

Banning all the paths to growth got the US to where it is now. Everything China did to out-compete the US was illegal in the US - the working conditions wouldn't have passed muster, the environmental damage wouldn't have been acceptable, the investment in cheap energy was off the agenda and the focus on heavy industry was broadly against the policy position which focused on financialisation and growing service industries.

We can argue until the cows come home which policies were the important ones (it is clear that some of what China does is counterproductive) but a key factor was legislative restrictions in the West. That is the opposite of the liberal part of liberal democracy. A key part of liberalism is giving people freedom to better their own lives.

In the wise words of Chairman Mao:

"Political power grows out of the barrel of a gun."

note that the parenthetical isn't entirely true. it's a very mathematical measure of living standards, and sure, one number is much bigger than the other, which does mean something, but how happy are you really doomscrolling various feeds about how the world is doomed? The wrong person's gonna get elected, you just know it. How happy can you really be knowing that to be the future? Or, hey, know anybody who's died from fentanyl recently?

There's a lot GDP doesn't capture. Hell, in terms of buying more things and driving the economy, you could look at it and say divorce is good for GDP since everybody is now buying twice as much stuff, not to mention all the lawyers bills. Not sure I'd argue that it's great for living standards though. so yes, let's compare GDP for what it measures, money number divided by people number, and see that our number is bigger, yay. But simplifying that to living standards is lazy shortcut thinking. If a FAANG engineer makes $400k/yr, but minimum wage is $15/hr or $30,000, their average pay is $215k, which sounds pretty great! The living standards for two someones making $215k are much nicer than for the one person making $30k.

How much do you trust your neighbors? How about those city/rural/suburb folk, especially the ones the extremists on the opposite side of the political spectrum? That contribute to a good living standard?

GDP isn't a good measure of living standards, it's a neat number to compare, and means something, but it doesn't define living standards.

If your average Chinese person lives in a comfortable, well-built apartment that costs 1/4th of what it would cost in the US, drives a car that's similarly built but cost $10k instead of $40k, and eats food thats the quarter of the price, could you convincingly state that he has the quarter of the actual wealth, even though the numbers would indicate that.

There is a line of evidence/thinking that says relations between democracies are a lot more stable. So if you make everyone a democracy there would be no wars and a lot more peace. And peace is what humanity (or each individual) aspires to be.

Unfortunately this may result in some degree of intervention or politicking that others may not like. Intervention, specifically, is forcing judgement onto others.

Regarding these elites, you could say it's a calling from whatever Christian god that tells them to spread whatever good they deem it to be. But it may just as well be hubris, or human nature - a superiority complex that comes from being at the top of the chain of 'merit' and status. Anyway, it's happened before. Even in the colonial times there were British who thought colonizing India was a 100% good thing for humanity.

> Freedom is better because it removes hurdles to more rapid evolutions of efficiency, which leads to better utilization of energy, bringing better material life for every human.

Anything taken to extremes is bad. The chinese that studied in America and decided to go back made such a decision when they saw things like school shootings and riots.

Anyway, China dominating chip and AI will likely happen soon. They don't really need Taiwan for that.

There's just too many of them, and there'll always be diminishing returns, so at some point what they do will be alright even if it is slightly behind.

1. The U.S. fails to protect Taiwan. 2. All allies lose confidence in the U.S. nuclear umbrella, and they create their own nuke programs. 3. Japan and S. Korea produce nukes nearly overnight. Lots of other countries follow. We now have a much worse nuclear arms race.

China needs only one reason to make the argument for taking control of Taiwan. That's national security. Taiwan is extremely important as a strategic island for defense of the mainland. For example, go look at a map of Taiwan airspace vs Chinese airspace. Honestly, I find it almost ridiculous that such a great power is boxed in so much by a tiny neighbor.

Personally, I think Taiwan needs to take one for the team and commit to reunification with China. That would cool things down significantly. Maybe the Chinese could even throw in some deals on borders to give the Philippines and other neighbors a break.

The US no longer being able to print money (without massive inflation) out of thin air would force it to be more fiscally responsible like many other countries have to be.

Take Zuckerberg for example. He went in front of congress and spouted as much anti-china rhetoric as he can. He obviously does not want competition from Chinese social media companies such as TikTok. He wants TikTok banned. Meta even admitted that it hired a PR agency to try to connect TikTok to CCP.

Take Elon for example. He definitely wants Trump in office because he knows if Chinese EVs ever make it to American shores, Tesla would have fierce competition. Trump wanting massive tariffs on Chinese goods matters little to Elon. Elon doesn’t care if he has to pay $5 for a Chinese good at Walmart instead of $2.50. Only commoners will suffer. But as long as Elon gets extremely high tariffs on Chinese EVs, he’ll be happy.

For Schmidt, it’s clear that he has big ownership stakes in American tech companies. Chinese tech companies are a real threat to them in competition.

Citation needed.

This narrative that the US are the good guys and they are the police of the world is the propaganda of the empire. It's coolaid.

I don't think the Chinese or whatever other empire comes will be better.

But this Pax Americana is just lies. Go tell Vietnam, Chile, East Timor, Afghanistan, Iraq, and so on about it.

That great violence unleashed in various wars wasn't actually necessary for realizing the goals of the countries involved is a very important history lesson which can prevent future violence.

(Similarly, one of the motivations for World Wars - who gets control of the colonies, seems futile in hindsight as the system of colonies was anyway dismantled.)

China has a decent amount of structural issues, the US just has to hold on until they explode without exploding itself.

Maybe you think that America and allies should have just stepped aside and let communist armies capture the various states, but geopolitics in real life is a little more complex than that.

This is nonsense, there are plenty of very leftist countries that USA didn't attack, so that provably wasn't the reason.

Communism is first and foremost anti democratic and authoritarian in these countries, that is the main issue. If you take Sweden that was very close to going towards worker owned economy in the 80's USA didn't do anything to them, same with rest of Europe. Sweden since then became much more right wing and moved away from those ideas, but at the time it looked very likely.

From most measures China has peaked. Its state capitalism has gotten it a housing bubble and investment in industries with not enough customers. Combine that with its falling birth rate and high youth unemployment it would seem that the failures, not successes, are on display.

I’m not. What do you recommend me to read, in order to come to that realization?

It probably wasn't the right tone for this audience though. But it's a style of rhetoric that's probably intentional.

I don't know what the results of combining all that data with an LLM will look like, but I think recording our experience to be used by an AI will only get better.

Also, sensors like HRMs or activity monitors could give it insights into your emotional state.

It could replicate your reactions.