hckrnws
Watched it a while ago. Made me seriously think about AI and what we should use it for. I feel like all the entertainment use cases (image and video gen) are a complete waste.
The chatbots and image editors are just a side-show. The real value is coming in e.g. chemistry (Alpha fold etc all), fusion research, weather prediction etc.
The real value is coming in warfare.
Right. More accurate predictions for meta-data based killings which as championed by US in their war on terror
Metadata based killings are most likely a huge improvement from the prior state of affairs
Yeah. Let the leaders assassinate each other with drone strikes instead of indiscriminately bombing whole cities as they used to.
what gov't in modern day would fall because the leader was assassinated? the next in line would just step up, and now have a pissed population that will be in favor of ratcheting up beyond assassinations.
Comment was deleted :(
None of that has reached the market yet. If it was up to the sciences alone, AI couldn't bear the weight of its own costs.
It also needs to be vertically integrated to make money, otherwise it's a handout to the materials science company. I can't see any of the AI companies stretching themselves that thin. So they give it away for goodwill or good PR.
That's not really true. Commercial weather prediction has reached the market, and a drug (sorry, can't find the new s link) that was found by AI-accelerated drug discovery is now in clinical testing
Why are images and video a complete waste? This makes no sense to me.
Right now the generators aren’t effective but they are definitely stepping stones to something better in the future.
If that future thing produces video, movies and pictures better than anything humanity can produce at a rate faster than we can produce things… how is that a waste?
It can arguably be bad for society but definitely not a waste.
Parent said "entertainment use cases" are a complete waste, not all uses of images and video. I don't agree, but do particularly find educational use cases of AI video are becoming compelling.
I help people turn wire rolling shelf racks into the base of their home studio, and AI can now create a "how to attach something to a wire shelf rack" without me having to do all the space and rack and equipment and lighting and video setup, and just use a prompt. It's not close to perfect yet, but it's becoming useful.
> particularly find educational use cases of AI video are becoming compelling.
compelling graphics take a long time to create. for education content creators, this can be too expensive as well. my high school physics teacher would hand draw figures on transparencies on an overhead projector. if he could have produced his drawings as animations cheap and fast using AI, it would have really brought his teaching style (he really tried to make it humorous) to another level. I think it would be effective for his audience.
imagine the stylized animations for things like the rebooted Cosmos, NOVA, or even 3Blue1Brown on YT. there is potential for small teams to punch above their weight class with genAI graphics
If AI can produce movies, video and art better aka “more entertaining” then humans than how is it a waste?
But it’s not. I think most can agree that there really has not been any real entertainment from genAI beyond novelty crap like seeing Lincoln pulling a nice track at a skate park. No one wants to watch genAI slop video, no one wants to listen to genAI video essays, most people do not want to read genAI blog posts. Music is a maybe, based on leaderboards, but it is not like we ever had a lack of music to listen to.
Let me phrase it a bit differently, then: AI generated cats in Ghibli style are a waste, we should definitely do less of that. I did not hold that opinion before the documentary
Education-style infographics and videos are OK.
I’m not even talking about this. Those cat videos are just stepping stones for academy award winning masterpieces of cinema like dune. All generated by AI on a click in one second.
I'm glad you're not the sole arbiter for what is wasteful and what isn't.
Just because you disagree does not make them wrong though
Comment was deleted :(
You might have said the same thing about GPUs for 20 years when they were mostly for games, before they turned out to be essential for AI. All the entertainment use cases were directly funding development of the next generation of computing all along.
Practical things are probably treating diseases and more abundance of physical goods. More speculative/sci-fi is merging in some form with AI and maybe immortality which I think is the more interesting bit.
DeepMind's new [edit: apparently now old] weather forecast model is similar in architecture to the toys that generate videos of horses addressing Congress or cats wearing sombreros. The technology moves forward and while some of the new applications are not important, other applications of the same technology may be important.
Is it really similar? I was under the impression it's a GNN of a (really dense) polyhedron, not a diffusion model
GenCast is a diffusion model, but it is not the "new" one like I said. Apparently there is another one. https://arxiv.org/pdf/2506.10772
[flagged]
reposting this from youtube comment
From 1:14:55-1:15:20, within the span of 25 seconds, the way Demis spoke about releasing all known sequences without a shred of doubt was so amazing to see. There wasn't a single second where he worried about the business side of it (profits, earnings, shareholders, investors) —he just knew it had to be open source for the betterment of the world. Gave me goosebumps. I watched that on repeat for more than 10 times.
My interpretation of that moment was that they had already decided to give away protein sequences as charity, it was just a decision of all as a bundle vs fielding individual requests (a 'service').
Still great of them to do, and as can be seen it's worth it as a marketing move.
Another way to interpret this (and I don't mean it pejoratively at all): Demis has been optimizing his chances for winning a nobel prize for quite some time now. Releasing the data increased that chance. He also would have been fairly certain that the commercial value of the predictions was fairly low (simply predicting structures accurately was never the rate-limiting step for downstream things like drug discovery). And that he and his team would have a commercial advantage by developing better proprietary models using them to make discoveries.
I think that's a rather conspiratorial way of framing it.
I think it's more about someone trying to do the most good that was possible at that time.
I doubt he cares much about prizes or money at this point.
Also since selling Deepmind to Google, it's Google's shareholder's money really.
I also noticed this as well. Actually went back and watched it several times. It's an incredible moment. I keep thinking, "if this moment is real, this is truly a special person."
In my experience all DeepMind content ends up being a puff piece for Dennis Hassabis. It's like his personal marketing engine lol.
Perhaps they need more advertising around the correct spelling of his name.
Is that a good thing or a bad thing? Demis is after all a co-founder and CEO.
Makes it seem that AI is a one-man show while also feeding the hype cycle
Greg Kohs and his team are brilliant. For example, the way it captured the emotional triumph of the AlphaFold achievement. And a lot of other things.
One of the smart choices was that it omitted a whole potential discussion about LLMs (VLMs) etc. and the fact that that part of the AI revolution was not invented in that group, and just showed them using/testing it.
One takeaway could be that you could be one of the world's most renowned AI geniuses and not invent the biggest breakthrough (like transformers). But also somewhat interesting is that even though he had been thinking about this for most of his life, the key technology (transformer-type architecture) was not invented until 2017. And they picked it up and adapted it within 3 years of it being invented.
Also I am wondering if John Jumper and/or other members of the should get a little bit more credit for adapting transformers into Evoformer.
I want to watch it, but at the same time, it’s basically going to be an advert for Google. I’m not sure if I can put up with the uncritical fluff.
I would love to see a real (ie outsider) filmmaker do this - eg an updated ‘Lo and behold’ by Werner Herzog
It was directed by Greg Kohs, who is a real filmmaker and does not work for Google.
Yeah I don’t mean to say they’re not a real filmmaker or untalented etc, I mean more the context they’re doing it. That they’ve chosen to cover this topic themselves, and that they would show critical angles of it and not just promo + hagiography
Are you saying this movie production wasn't paid for by Google? If it was, surely he did?
oh it might have been paid for by Google for sure.
Like 99.99% probability, sure. Greg's previous big feature was on Deepmind's AlphaGo, three years after its Google acquisition.
Full length: https://www.youtube.com/watch?v=WXuK6gekU1Y
They do a great job capturing the "Move 37" moment: https://youtu.be/WXuK6gekU1Y?t=2993
It's an advert for Demis Hassabis, not Google.
Where he speaks french
Watched it this week. Pretty good.
There are a couple parts at the start and the end where a lady points her phone camera at stuff and asks an AI about what it sees. Must have been mind-blowing stuff when this section was recorded (2023), but now it's just the bare minimum people expect of their phones.
Crazy times we're living in.
It's official too. It's on: https://www.youtube.com/@googledeepmind
Moderators: Please change the link; feels kind of unethical to bait someone into paying for this now.
i tried to watch it but like AI in general, it was extraordinarily boring. neural nets are really cool technically, but the whole AI thing is just getting old and I couldnt care less where its going
we can guarantee that whether its the birth of superintelligence or just a very powerful but fundamentally limited algorithm, it will not be used for the betterment of mankind, it will be exploited by the few at the top at the expense of the masses
because thats apparently who we are as a species
Hi, I’m genuinely curious about your writing style. I’m seeing this trend of no proper casing and no punctuation becoming vogue-ish. Is there a particular reason you prefer to write this way or is this writing style typical for a generation? Sincere question, not snark, coming from an older generation guy.
This is the writing style of this generation. I've just scrolled 6 months of my conversation with a friend in his twenties. Not a single comma or period to be seen. I mean on his side.
If you grew up in the internet of early 2000s, that's how we wrote online.
I grew up in the Internet at that time, and it's certainly not how I type. So you might want to be more specific about which sites or subcultures you think this style is representative of?
I’m certainly no authority but i tend to write the same way for casual communication, came from the 90s era BBS days. It was (and still is) common on irc nets too. Autocorrect fixes up some of it, but sometimes i just have ideas i’m trying to dump out of my head and the shift key isn’t helping that go faster. Emails at work get more attention, but bullshittin with friends on the PC? No need.
I’ll code switch depending on the venue, on HN i mostly Serious Post so my post history might demonstrate more care for the language than somewhere i consider more causal.
Comment was deleted :(
If you watch on there's a bit where they decide to give away all the protein folding results for free when they could have charged (https://youtu.be/d95J8yzvjbQ?t=4497). Not everything is exploitation rather than the betterment of mankind.
Correct! I’m glad people are finally starting to get it
weekends are always better on hn
How does that relate to my comment?
Just watched it yesterday and enjoyed every second of it, the director put more focus on Demis Hassabis which turns out to be a true superhero and I have to confess that I am probably admiring him more that any other human in the tech industry.
I caught it on the airplane a few days ago. I would have loved a little more technical depth, but I guess that's pretty much standard for a puff piece.
It is interesting that Hassabis has had the same goal for almost 20 years now. He has a decent chance of hitting it too.
I find it funny that the YouTube link takes you to the film, but like an hour into it.
Yes, it made me think I'd already watched it and had forgotten about it...
Loved this documentary. People complaining - WTFV first.
Streaming on YouTube now: https://www.youtube.com/watch?v=d95J8yzvjbQ
Hard to discount the impact of AlphaFold in science work but submitting this to a number of film festivals like Tribeca seems a bit AI-washing.
What is AI-washing?
AlphaFold is optimization, not thinking. Propaganda 'r us.
Sure, but AlphaFold is still probably the most impactful and positive thing to have come out of "Deep Learning" so far.
Did you watch the documentary? Would probably fare better if you did, because it'd give you the context for the film title.
I'm an hour into it, unconvinced.
The illusion that agency 'emerges' from rules like games, is fundamentally absurd.
This is the foundational illusion of mechanics. It's UFOlogy not science.
Well, two things: it's the last sentence of the film; being on hour into something you're calling propaganda is brave.
Anyways. I thought the documentary was inspiring. Deepmind are the only lab that has historically prioritized science over consumer-facing product (that's changing now, however). I think their work with AlphaFold is commendable.
It's science under the creative boundary of binary/symbols. And as analog thinkers, we should be developing far greater tools than these glass ceilings. And yes, having finished the film, it's far more propagandic than it began as.
Science is exceeding the envelop of paradox, and what I see here is obeying the envelope in order to justify the binary as a path to AGI. It's not a path. The symbol is a bottleneck.
Everything between your ears is an electrochemical process. It's all math and there is no "creative boundary." There's plenty to criticize in AI hype that we're going to get to machine intelligence very soon. I suspect a lot of the hype is oriented towards getting favorable treatment from the government if not outright subsidies. But claiming that there are fundamental barriers is a losing bet.
It doesn't happen "btwn ears" and math is an illusion of imprecision. The fundamental barrier is frameworks and computers will not be involved. There will be software obviously. But it will never be computed.
Plenty *commercial* labs frequently prioritized pure science over *immediate* consumer products, but none done so out of charity. Deepmind included.
Your mind emerges from a network of neurons. Machine models are probably far from enabling that kind of emergence, but if what's going on between our ears isn't computation, it's magic.
It's not magic. It's neural syntax. And nothing trapped by computation is occurring. It's not a model, it is the world as actions.
The computer is a hand-me-down tool under evolution's glass ceiling. This should be obvious: binary, symbols, metaphors. These are toys (ie they are models), and humans are in our adolescent stage using these toys.
Only analog correlation gets us to agency and thought.
Comment was deleted :(
Is there a fundamental difference between it and true agency/thought? I’m not so sure.
Agency will emerge from exceeding the bottleneck of evolution's hand-me-down tools: binary, symbols, metaphors. As long as these unconscious sportscasters for thought "explain" to us what thought "is", we are trapped. DeepMind is simply another circular hamster wheel of evolution. Just look at the status-propaganda the film heightens in order to justify the magic.
Why is it absurd? Because believing that would break some deep delusion humans have about themselves?
Not sure why this is downvoted. The comment cuts to the core of the "Intelligence vs. Curve-Fitting" debate. From my humble perspective as a PhD in the molecular biology /biophysics field you are fundamentally correct: AlphaFold is optimization (curve-fitting), not thinking. But calling it "propaganda" might be a slight oversimplification of why that optimization is useful. If you ask AlphaFold to predict a protein that violates the laws of physics (e.g. a designed sequence with impossible steric clashes), it will sometimes still confidently predict a folded structure because it is optimizing for "looking like a protein", not for "obeying physics". The "Propaganda" label likely comes from DeepMind's marketing, which uses words like "Solved"; instead, DeepMind found a way to bypass the protein folding problem.
If there's one thing I wish DeepMind did less of, it's conflating the protein folding problem with static structure prediction. The former is a grand challenge problem that remains 'unsolved' while the latter is an impressive achievment that really is optimization using a huge collection of prior knowledge. I've told John Moult, the organizer of CASP this (I used to "compete" in these things), and I think most people know he's overstating the significance of static structure prediction.
Also, solving the protein folding problem (or getting to 100% accuracy on structure prediction) would not really move the needle in terms of curing diseases. These sorts of simplifications are great if you're trying to inspire students into a field of science, but get in the way when you are actually trying to rationally allocate a research budget for drug discovery.
I think if you watch the actual film you'd find they don't claim AlphaFold is thinking.
I'm concerned that coders and the general public will confuse optimization with intelligence. That's the nature of propaganda, substituting sleight of hand to create a false narrative.
btw an excellent explanation, thank you.
What's the difference between optimisation and intelligence?
what is thinking?
Sharp wave ripples, nested oscillations, cohering at action-syntax. The brain is "about actions" and lacks representations.
Creatively peeling the hyper dimensional space in the scope of simplectic geometry, markhov blanket and helmholtz invariance????
Crafted by Rajat
Source Code