Comment by phtrivier
1 year ago
Okay, I'm really in a sad mood, so: tell me there will be places like that, again, somewhere, ever ?
We need this. Like, really, we need someone to have created the xerox part of the 21st century, somewhere about 20 years ago.
I honestly though Google would be that - but apparently it's easier to fund R&D on "selling copying machines" than "selling ads". Maybe "selling ads" earn _too much_ money ? I don't know.
I know, I know, DeepMind and OpenAI and xAI are supposed to fix climate change any time soon, and cure cancer while they invent cold fusion etc, etc... and it's only because I'm a pessimistic myopist that I can only see them writing fake essays and generating spam, bad me.
Still. Assuming I'm really grumpy and want to talk about people doing research that affects the physical world in positive way - who's doing that on the scale of PARC or Bell ?
The secret hero of that time was the US government. I’m not talking about the MIC, which is still quite robust and more bad than good. I am speaking more broadly. If you had a practical PhD and were willing to show up at a place at 9:00, you could get a solid upper middle class job with the Feds where you couldn’t get fired unless you broke the law.
The government also has always kept academia afloat. It is a privilege afforded to professors to believe they do not work for the state, but they do.
Great government and academic jobs forced companies to create these labs where it was better to hire great people and “lose” some hours to them doing whatever they want (which was still often profitable enough) than have zero great people. Can you imagine Claude Shannon putting up with the stuff software engineers deal with today?
The other main change is that how to run big companies has been figured out well enough that “zero great people” is no longer a real survival issue for companies. In the 1970s you needed a research level of talent but most companies today don’t.
Something that just dawned on me is the downstream effects of United States’ policy regarding science during WWII and the Cold War. The Manhattan Project, NASA, the NSA and all of its contributions to mathematics and cryptography, ARPA, DARPA, and many other agencies and programs not only directly contributed to science, but they also helped form a scientific culture that affected not only government-ran and government-funded labs, but also private-sector labs, as people and ideas were exchanged throughout the years. It is a well-documented fact that Xerox PARC’s 1970’s culture was heavily influenced by ARPA’s 1960’s culture.
One of the things that has changed since the 1990s is the ending of the Cold War. The federal government still has national laboratories, DARPA, NASA, the NSF, etc. However, the general culture has changed. It’s not that technology isn’t revered; far from it. It’s just that “stopping Hitler,” “beating the Soviets,” and grand visions for society have been replaced with visions of creating lucrative businesses. I don’t hear about the Oppenheimers and von Neumanns of today’s world, but I hear plenty about Elon Musk and Sam Altman, not to disrespect what they have done (especially with the adoption of EVs and generative AI, respectively), but the latter names are successful businessmen, while the former names are successful scientists.
I don’t know what government labs are like, but I know that academia these days have high publication and fundraising pressures that inhibit curiosity-driven research, and I also know that industry these days is beholden to short-term results and pleasing shareholders, sometimes at the expense of the long-term and of society at large.
I don’t hear about the Oppenheimers and von Neumanns of today’s world
Sadder still is the underlying situaiton behind this: the fact that there's nothing of even remotely comparable significance happening in the public sphere for such minds to devote themselves to, as those man did. Even though the current civilization risk if anything significantly greater than in their time.
9 replies →
> It’s not that technology isn’t revered; far from it. It’s just that “stopping Hitler,” “beating the Soviets,” and grand visions for society have been replaced with visions of creating lucrative businesses.
Any kind of societal grand vision we had has been falling apart since about 1991. Slowly at first (all the talk about what to do with the "peace dividend" we were going to get after the fall of the Soviet Union) And that accelerated with the advent of the internet and then accelerated even more when social media came on the scene. We no longer have any kind of cohesive vision for what the future should look like and I don't see one emerging any time soon. We can't even agree on what's true anymore.
> I don’t know what government labs are like
Many of these are going to be in danger in the next administration especially if the DOGE guys get their way.
1 reply →
> successful businessmen, while the former names are successful scientists
We’ve seen this before with Thomas Edison.
>It’s not that technology isn’t revered; far from it. It’s just that “stopping Hitler,” “beating the Soviets,” and grand visions for society have been replaced with visions of creating lucrative businesses
Universities are tripping over themselves to create commercialization departments and every other faculty member in departments that can make money (like CS) has a private company on the side. Weird that when these things hit, though, the money never comes back to the schools
1 reply →
Yup. Silicon Valley would not exist without large government spending.
You can bet this spending is going to be among the fist things slashed by DOGE-lile efforts ("Scientists ? They're just liberal elites wasting our hard earned money researching vaccines that will change your dog's gender in order to feed it to communist immigrants.")
I suppose I could be cheered up by the irony, but, not today.
> honestly though Google would be that - but apparently it's easier to fund R&D on "selling copying machines" than "selling ads". Maybe "selling ads" earn _too much_ money ? I don't know.
I'm pretty sure Google Brain was exactly what you are looking for: People like to think of DeepMind, but honestly, Brain pretty much had Bell Labs/PARCs strategy: they hired a bunch of brilliant people and told them to just "research whatever is you think is cool". And think all the AI innovations that came out of Brain and were given to the world for free: Transformers, Vision Transformers, Diffusion Models, BERT (I'd consider that the first public LLM), Adam, and a gazillion of other cool stuff I can't think of right now.... Essentially, all of the current AI/LLM craze started at Brain.
Yes, it was basic research( guided to the field of machine learning), but between a search monopoly and their autonomous car project, they definitely have a great economic engine to use that basic research and the talent it pulled into Google, even if a lot of it escaped.
Right. And I'm sure that if I ever get in a better mood, I'll find that the current AI/LLM craze is good for _something_.
Right now the world needs GWh batteries made of salt, cheap fusion from trash, telepathy, a cure for cancer and a vaccine for the common cold - but in the meantime, advertisers can generate photos for their ads, which is, _good_, I guess ?
Your problem stems from assuming our natural state is some Star Trek utopia, and only our distraction by paraphernalia is preventing us from reaching such a place. Like we are temporarily (temporally?) embarrassed ascended beings.
Humanity’s natural state is abject poverty and strife. Look at any wealth graph of human history and note how people are destitute right up until the Industrial Revolution, and then the graph explodes upward.
In a way we (well, especially the West) are already living in utopia. You’re completely right that we can still vastly improve, but look back at the progress we already made!
It does sound like you're in a particularly bad mood, so yes, maybe our outlook does change. Maybe it helps to think of a darker timeline where Google would have kept all of these advances to itself and improved its ad revenue. Instead it shared the research freely with the world. And call me naive, but I use LLMs almost daily, so there definitely _is_ something of value that came out of all this progress. But YMMV, of course.
2 replies →
Can't you get telepathy from training AI on functional MRI data? And then finding a way to pinpoint and activate brain regions remotely?
I mean brain-machine interfaces have been improving for quite a while.
Telepathy might even already exist.
2 replies →
Rolling back the 1980s neoliberal cultural ideals of letting markets and profits be the highest arbiter of societal direction is the key.
Silicon Valley hippies have been replaced by folks focussed on monetisation and growth.
It’s not great for the west, but those problems are being tackled. We just don’t get to read about it because ‘China bad’ and the fear of what capital flight might do to arguably inflated US stock prices
https://www.energy-storage.news/byd-launches-sodium-ion-grid...
Once we get superintelligence — some time next year I’d say — then we will have a tool to make all those dreams come true.
4 replies →
Extreme ultraviolet lithography originated with paper out of Bell Labs in 1991, then US government funded multiple research efforts via national nuclear research labs that came up with a potential method to implement but it took 20 more years of trial and error by ASML to make a practical machine they could sell. Other companies tried and gave up because of the technical challenges. This advance is responsible for modern chip fabs fastest chips.
No, but you also shouldn't romanticize Bell Labs _too_ much. It was not exactly a fun place to work. You got a 2 year postdoc and then you were out, and those two years were absolutely brutal. Its existence was effectively an accounting fluke. Nothing like it really exists now because it would largely be seen as an inefficiency. Blame Jack Welch, McKinsey, KKR, HBS or whoever you like.
I was a postdoc there and I would not say it was brutal. I got a very good salary (far above an academic postdoc), health benefits, relocation, the ok to spend $1000/day on equipment with no managerial review[0], and access to anyone and everyone to whom I felt like speaking. I read horror stories in Science and other journals about people's experiences elsewhere and am grateful that I was spared so much nonsense. It was the greatest university I have ever set foot in. I still feel unworthy of the place.
[0]This was in the early '90s when $1000 went a long way.
> It was not exactly a fun place to work.
I couldn't disagree more, but perhaps the time I was there (late 90s) was different.
Just curious, are you speaking from personal experience?
Hmm
There are companies that push many various technologies
Samsung conglomerate does everything, Intel does hard (semiconductor research, manufacturing) and soft (computer science/software) things
Maybe we're at the point where you need to specialize in one industry, so achieving various stuff like they did at Bell is harder?
The answer is Google. I'm not sure why you're being so dismissive.
When we look back in 20 years, things like the transformer architecture, AlphaFold (which just won a Nobel prize), and Waymo are going to have improved the world in a positive way as much as anything Bell Labs did, and certainly more than PARC.
>I honestly though Google would be that - but apparently it's easier to fund R&D on "selling copying machines" than "selling ads". Maybe "selling ads" earn _too much_ money ? I don't know.
Google has put quite a bit of resources into quantum computing research. It's not just for selling ads, though I have no doubt that it will be used for that among other things. But right now there's still no guarantee it's going to pay off at all.
[dead]