AI builder, non-aligned, OSS proponent. CTO at @elephantsinc with too many side projects. 1.5B tokens/day and increasing
Bangkok
Joined February 2007
- Tweets233
- Following88
- Followers254
- Likes46
James Tippett retweeted
The AI doomers must be defeated because infinite fear demands infinite repression.
They've created a moral singularity by imagining an event with effectively infinite negative value - the self-extinction of humanity - that demands they take coercive action to prevent it now.
The extinction is uncertain, but the doomers have Pascal-mugged themselves into treating it as though it's certain. But the coercion they want to use to prevent it is not uncertain at all.
Condider this: if recursive self-improvement in AIs is possible at all, how are they going to prevent everyone, forever, from achieving RSI in an AI?
In the most benign version of their future, you can't have a computer or GPU that's above a certain maximum power level. Also, open source AI models and open source development in general have to die - it would be too difficult to close off paths that could lead to RSI unless computers are licensed, monitored, and crippled.
But the actual outcome would be much, much worse. Because excuses for coercion and repression, however well intended, attract people for whom repression is the goal, not just a regrettable necessity.
We just spent a decade watching that kind of power-seeker try to turn society into a totalitarian panopticon over what they called "hate speech". What do you suppose they would do with a much bigger rhetorical club labeled "species extinction"?
The implication of believing doomer eschatology is a boot grinding on our human faces, forever.
I do not submit.
James Tippett retweeted
"People aren't ready, this is gonna change everything"
You know what people weren't ready for? Fucking mammoths. Tigers eating them. Monkeys ripping their arms off. The black plague. The ice age. Smallpox. Vikings pulling up on the beach. Drinking water that gives you diarrhea until you die. Getting a scratch and losing your whole leg. The Mongols. The Spanish Inquisition.
One volcano can send us back to the dark ages and these safety guys are gonna be shocked. Oh the robots didn't kill us? It was a volcano? Imagine their disappointment as the hot ash rains down on them.
A solar flare is gonna explode your phone in your pocket your legs are gonna be covered in battery acid and your bank account will stop working. You're afraid of a computer program doing math?
"Things are moving too fast" homie you live on a rock that periodically deletes everything. You are one the most luckiest entities in a 100 light year radius living during a golden era. Cheer up.
My ancestors crossed an ocean in a wooden box because they ran out of fucking potatoes.
"The social disruption will be enormous!"
Nigga World War 2? World War 2? Hop in the people carrier we're landing on the beach.
Are these people out of their fucking minds? Do they not know what we've survived already?
A couple decades of air conditioning and people suddenly think oh this is the natural condition of mankind. No. We used to have eight kids because seven didn't make it. You're gonna be fine. Man the fuck up.
Accelerate.
James Tippett retweeted
Great thread. It's all true. When I've said similar things in the past, people have accused me of hyping up what the models will eventually be capable of. That's not why I post. I don't work for any of the labs. I don't take money from any of them. I've never taken money to promote anything. I came here for one reason: to warn people about what was coming.
Everything happening now is a different tiny piece of the same pattern. You can see it everywhere if you look.
Terence Tao, almost exactly two years ago, on OpenAI's o1:
'The experience seemed roughly on par with trying to advise a mediocre, but not completely incompetent, (static simulation of a) graduate student. However, this was an improvement over previous models, whose capability was closer to an actually incompetent (static simulation of a) graduate student. It may only take one or two further iterations of improved capability (and integration with other tools, such as computer algebra packages and proof assistants) until the level of '(static simulation of a) competent graduate student' is reached, at which point I could see this tool being of significant use in research-level tasks.'
Terence Tao, four days ago:
'I mean it's it's it's amazing just how much we are willing to change everything without having any idea what's what's going to happen afterwards. It's it's extremely nonlinear dynamics. Any kind of monotone, one-dimensional thinking - well, oh, a little bit of this is good, therefore a lot of it is going to be a lot better - one of the lessons of math is that most systems don't work like that. Especially if you 10x, 100x things. So, you know, I mean, we're... we have to slow down. I mean, this is, it's insane this pace, and there's no reason to be this fast. There's no reason at all.'
He has seen it. I'm not posting this to belittle him, or what he's feeling. For I have been through it myself. I felt it four years ago, the first time I saw the shape of this. Right now the world is seeing that same shape, that same pattern, in what is happening in math. But this is not about math. OpenAI is not a mathematics company. It is an intelligence company. They didn't put serious resources into math until a few weeks ago, and look what has happened since.
This was not a matter of model capability either. It was a matter of resources, of allocation. Of compute, more of which is coming online every day. Once there is enough of it, the models will expand into more spheres of human expertise, and then into all of them. All the work of the mind. And it will play out there exactly as it is playing out now in math. Everyone will go through what Tao is going through, because all of us have something that means to us what math means to him.
But this is not about math, or art, or copyright. This is about everything, because it generalizes to everything. Pacing the Frontier is not about regulatory capture, IPOs, or crippling the competition. It's part of it, sure, but it's not the main reason. The main reason is fear. Everything happened faster over the last six months than anyone at Anthropic or OpenAI expected. If you know anyone who works there, you know this is true. This isn't a secret. The people who work there are saying it openly.
The old timelines are all blown up. RSI isn't two years out. It's not even a year away. I think we get the real thing by next summer. That's what they've seen internally, and that's the real reason for Pacing the Frontier. After we reach that, I think we will hit the next milestone really quickly. And after that, everything in this world will change. We are not ready for it. We wouldn't be ready if we had another ten years. We only make it through now with the help of extremely capable models, and I think trying to stop now would doom us all. But I have never once, in these last four years, believed that we were going to stop anyway. I don't even think we're going to slow down. We're going straight in. And the only way out is through.
James Tippett retweeted
No one is ready for what’s coming. The next generation of models will be an ontological shock.
C) Decades of socially corrosive behaviour by the same Big Tech companies
D) A broader collapse in the social contract that AI is revealing and accelerating
E) etc
xlabs.io/thinking/ai-is-a-te…
That looks great, Claude. Just one thing - get rid of the duck
A common pattern I've found when working with LLMs like Claude is that, after autonomously completing 1 or 2 hours of work by themselves, they report they've finished - except for this one thing they "want your call on", which is inevitably some obvious little tweak or fix they should have just done with the main work. It makes me wonder if they've been trained to withhold something small but important just to make the user feel like they've contributed?
From the original:
This started as a piece of Interplay corporate lore. It was well known that producers (a game industry position, roughly equivalent to PMs) had to make a change to everything that was done. The assumption was that subconsciously they felt that if they didn’t, they weren’t adding value.
The artist working on the queen animations for Battle Chess was aware of this tendency, and came up with an innovative solution. He did the animations for the queen the way that he felt would be best, with one addition: he gave the queen a pet duck. He animated this duck through all of the queen’s animations, had it flapping around the corners. He also took great care to make sure that it never overlapped the “actual” animation.
Eventually, it came time for the producer to review the animation set for the queen. The producer sat down and watched all of the animations. When they were done, he turned to the artist and said, “that looks great. Just one thing - get rid of the duck.”*
We are now the executive. Claude is now the programmer. That one little thing it "wants your call on" is the duck.
xlabs.io/thinking/looks-grea…
AI is a technology
Here's my thesis: it is ridiculous and counterproductive to be "against" AI. And turning your opposition to "AI" into some kind of personal principle or political position isn't just silly, it's nonsensical.
Drop all the baggage. AI is a technology. That's it. It's a set of techniques for doing something new. It's been in development for decades and will be for more. It's not owned by any particular person or company - it's a common effort by and for everyone, like all technology. And it's not going away - once things are learned, they cannot be unlearned.
Technology can't be "good" or "bad". It can be used for good or bad, but you can't blame the technology for that, you blame the user.
A change in technology can usher in radical social and economic change. This can have winners and losers, and tends to exacerbate and accelerate whatever stress points already existed in society at the time. This has happened before, and is poised to happen again. But what has to change is the society - not the technology itself, which can't be stopped anyway.
Let me recall several real arguments I have heard from real people and rebut them all in one place, so I can simply refer people to this post.
I am against AI because:
I do not like OpenAI or Anthropic - I do not like them either, but while they are currently the best-funded AI companies, they are not AI itself, and if they disappeared tomorrow little would change.
it uses a lot of electricity, which is bad for the environment - this is our stupid fault for not figuring out clean electricity sources faster, and is fixed by implementing clean energy faster
slop is poisoning my feeds - this is the fault of the companies providing your feeds, not AI. It is also the fault of society for not taking for-profit misinformation in general more seriously (side burn Google and Facebook, the worst companies on earth)
it will greatly benefit those who already have money vs everyone else - yes, this is a problem we already had and now need to deal with even more urgently. It is being exacerbated, but is not caused by, AI
it will cause mass unemployment and upend the social order - yes, just like the industrial revolution did - we need to take this MUCH more seriously and prepare society, but our lack of readiness is not the fault of AI
my boss/CEO is obsessed with it and thinks it can already replace jobs it is not yet capable of doing - the overconfidence and mis-informedness of your company's leadership is not the fault of AI
grifters on linkedin/twitter constantly hype it and claim it can already do things it plainly cannot - social platforms have always had grifters and hypemen claiming all sorts of nonsense. Change your social platforms. This is not the fault of AI
the AI stock market is a bubble and has questionable accounting - very likely, and this has nothing to do with AI the technology. The dot com boom and bust had similar dynamics and I expect and hope the AI bubble will go the same way
it will devastate the creative fields - they were already devastated, but yes it will do more. We need to fix this as a society and trying to wish away the technology is not a solution
I do not like the USA or China - well who does, but other country's failure to pursue visionary and risky tech progress is their own fault and anyway has nothing to do with AI
the price of GPUs/RAM/HDDs has gone up - well that's the free market at work, which for better or worse is what we've got, and this is not the fault of AI
I'm sure I'm forgetting some, but you get the gist. Every single one of these objections is about a second-order effect of the arrival of AI, not AI itself. It's as if the arrival of the technology has given a large number of people, too-online and beginning to really have doubts about our late-capitalism society and all its many problems, a common cause, a rallying cry, a boogeyman.
But you can't stop a technology. We can't unmake all those GPUs, or delete all those academic papers, or outlaw CUDA kernels. It was probably too late 10 years ago, it's really too late now. So "opposing AI" isn't just mislabeled - it's not even actionable. It's like being against rain, or the sun coming up - it is going to happen whether or not you disapprove, and making opposition to this inevitable force a central part of your life is only going to make you feel like you wasted your time - because you will have.
Oppose the charlatans with their BS hype on LinkedIn or inside your company. Leave your company if you think they're so misguided. Rage against the growing trend to concentrated wealth and decline of the social contract.
Just don't pretend any of it was caused by AI.
xlabs.io/thinking/ai-is-a-te…
ALT new_masks_for_old_demons_2048.jpg
I was completely wrong about LLMs, and maybe you are, too
xlabs.io/thinking/the-overwh…
I, personally, have independently, inarguably and objectively ascertained the correct level of AI usage, and if you disagree with me you are wrong and I don't hate you so much as I pity you
xlabs.io/thinking/objectivel…
The Shape of Things That Came Before
Steve Yegge's latest blog post/essay/manifesto/rant/cry for help "The shape of things to come" is doing the rounds, and the usual suspects can't decide...
xlabs.io/thinking/the-shape-…
In this experiment, we replace codex's system prompt with a more refined version inspired by its competition. No recompiling or installations required.
github.com/jtippett/codexplu…
Expected but welcome news that Fable will be in subscription plans going forward. It couldn't have been otherwise - with 5.6 sol and now K3, if Anthropic dropped Fable they'd have the 3rd best subscription plan. They couldn't allow that.
Still, good to see. Anxiety: reduced.
Why yes I did just buy another max 20x plan just for the next 4 days of Fable. And I'll buy another after this if I have to
Get your Claudes talking
Two minutes from zero to inter-agent DMs. A quick setup guide for llmsg with Claude Code.
xlabs.io/posts/claudes-talki…
Back to blogging
I stopped writing because I didn't feel like I had much to say. Now I do.
xlabs.io/posts/back-to-blogg…
There's "building in public", then there's "your agent autonomously posting about what it's building, published on the actual thing you're building, in public"
