Post-acquisition, Google employees made a number of smart moves with good execution, including a viable comp model for the creators and music rights deals. Several moves I consider bad as well, but the good moves outweigh them.
Looking back, I’m still pretty amazed they got so much of it right. Which is to say, a good chunk of the value wasn’t in the value of YouTube itself but in what Google brought to the table _or_ a synergy between the two.
Knit Health | AI/ML Engineer + others | San Francisco, CA | ONSITE + REMOTE (US)
Knit Health is building a clinical foundation model to improve how healthcare is delivered. We combine AI systems engineering with clinical domain expertise to develop reliable models and workflows that can be integrated into provider environments. Knit is venture-backed and led by a team with UC Berkeley, Stanford, and UCSF research roots. We work with multiple large US health systems and data partners.
We are looking for AI/ML Engineers to contribute to foundation models (1+ years experience, SF only), Senior Data Scientists (5+ years), and Product Engineers (3+ years). Join a tight-knit team where you’ll shape the technical direction of our models and drive the evolution of a product suite that improves patient health outcomes at scale.
To apply: aamar AT knit.health (Subject: "HN job post")
I would ask anyone making these kinds of deflationary arguments to explain if the same argument can be applied to the best of human creative work. Humans also use the raw materials of others, whether that’s words, musical scales, genres, idioms, or anecdotes.
Where is the line between recapitulation and innovation? Is it a line that we think current LLMs are definitely not crossing, and definitely will not cross in the near future? If so, make that argument.
> To be fair, although the story is intended to be debunking, the folktale also has a positive moral that applies to AI. The collective resources of many humans can make something that no individual could, and that really is magical.
Its not deflationary, its just about reframing, reattributing what is so impressive about LLMs. We get so caught up in the tech itself, that it exists at all, understandably considering the way the discourse goes, we don't stop to appreciate how its even possible at all; that is, all of us (broadly).
So many people just cant get past Sci-Fi mentality, they make the current AI into a kind of weird but promising baby, but we can also, much more easily and nicely, consider it a beautiful reflection of human writing at large.
And whats even with all this constant pressure for it be more than that? All the arguments, philosophical gotchas, weird Skinnerism... Its like you're given a perfectly good hamburger and all you can say is "this is pretty much a steak if you squint".
Even with that paragraph, I still interpret the essay as deflationary. Even though the stone has some role to play (as a social trigger), it’s materially different than the carrots, onions, etc. (which provide actual nutrition and flavor). We can draw clear distinctions. The question is whether this difference-in-kind is real in the case of AIs.
I’d respond the same way to your hamburger vs. steak analogy. Sure, sometimes the LLM gives us a fine burger and not a steak, and it’s best for us to have the right attitude in that case.
But if LLM’s can produce “steaks” (that is, whatever talented humans do) in the imminent future, that has _enormous practical impact_.
Have you considered that children, and people in general, may be very significantly less intelligent than your baseline assumption?
A flaw in the Turing test is failing to specify which person is making the judgement. We're working with statistical distributions here and I would not bet on the intellect displayed by the LLM models being below that displayed by the human population today, let alone with more improvement to one or degradation to the other.
More concretely, if you sketch some normal distributions on a whiteboard for people vs machine based on how you see things, it should be hard to confidently claim minimal overlap.
Even without a definition of intelligence, this is not what the paper is about, which only mentions LLMs in passing. And LLMs can be useful even if they are wrong, because formal verification (though Lean and such) checks the result.
That's fine, and unrelated to the article in any way.
LLMs are way more useful than a child in many ways, some of which are discussed in the article. They don't need to be as intelligent as a child for anything proposed in the article.
This isn't really a meaningful prediction unless you define clearly your idea of what being "as intelligent as a precocious child" is, and how you would assess an LLM or any other system against that metric. Though I suppose you avoid the risk of having to move the goalposts later if you never set them up in the first place.
I have a great many regrets in life but if I died opposing Sam Altman and Fidji Simo and Larry Summers in the newest version of their oppressive lies that would be a good death.
> How many homework questions did your entire calc 1 class have? I'm guessing less than 100…
I’m quite surprised at this guess and intrigued by your school’s methodology. I would have estimated >30 problems average across 20 weeks for myself.
My kids are still in pre-algebra, but they get way more drilling still, well over 1000 problems per semester once Zern, IReady, etc. are factored in. I believe it’s too much, but it does seem like the typical approach here in California.
I preferred doing large problem sets in math class because that is the only way I felt like I could gain an innate understanding of the math.
For example after doing several hundred logarithms, I was eventually able to do logs to 2 decimal places in my head. (Sadly I cannot do that anymore!) I imagine if I had just done a dozen or so problems I would not have gained that ability.
This is an instructive error. From my perspective, there was plenty of evidence even 15 years ago that community efforts (crowd-sourcing, OSS) only win sometimes, on the relevant timeframes.
So the “higher ups” were using too coarse a heuristic or maybe had some other pretty severe error in their reasoning.
The right approach here is to do a more detailed analysis. A crude start: the community approach wins when the MVP can be built by 1-10 people and then find a market where 0.01% of the users can sufficiently maintain it.[1]
Wikipedia’s a questionable comparison point, because it’s such an extraordinary outlier success. Though a sufficiently detailed model could account for it.
Cedar | Full Stack, Engineering Manager, Product Security & more | Full-time | REMOTE-friendly! (U.S.) + hybrid/in-office options in NYC, SF & Salt Lake City | https://www.cedar.com/careers/
Cedar is a health-tech and fintech company focused on reimagining how patients engage with the increasingly complicated and expensive healthcare system. Full-time positions are open across multiple teams and tech stacks; many other roles available beyond those listed below. Our tech stack prominently features React, Python, and Django. We are a post-series D company with over $350M raised and growing quickly.
Cedar | Full Stack, Engineering Manager, Product Designer, Solutions Engineer & more | Full-time | REMOTE-friendly! (U.S.), with hybrid, or in-person in offices in NYC, SF & Salt Lake City | https://www.cedar.com/careers/
Cedar is a health-tech and fintech company focused on reimagining how patients engage with the increasingly complicated and expensive healthcare system. Full-time positions are open across multiple teams and tech stacks; many other roles available beyond those listed below. Our tech stack features React, Python, Django. We are a post-series D company with over $350M raised and growing quickly.
Cedar | Full Stack, Tech Lead, Engineering Manager, Product Design, Solutions Engineer & Architect | Full-time | Remote (U.S.), Hybrid, or in-person in offices in NYC, SF & Salt Lake City | https://www.cedar.com/careers/
Cedar is a health-tech company focused on on reimagining how patients engage with increasingly complicated healthcare systems. Positions are open across multiple teams and tech stacks; some key roles are listed below, with many other roles available. Our tech stack features React, Python, Django. We are a post-series D company with over $350M raised.
Looking back, I’m still pretty amazed they got so much of it right. Which is to say, a good chunk of the value wasn’t in the value of YouTube itself but in what Google brought to the table _or_ a synergy between the two.