Then what are we talking about? If this is only about comfort of people who don't give a shit anyways then just relate it to money or cartoons or whatever and walk away. It just doesn't matter.
Some of the exams in Berkeley were brutal, but they never felt like trick questions, they did on occasion require a level of mastery of the material which was extreme, but it never felt like someone was just trying to make the questions obtuse for the sake of it.
No one consented to training llms, as the op clearly implies, if they had been asked they would have declined to do so. As would all of the many copyright holders who are in the process of suing the model companies.
S-1 isn’t public yet. Source on the lockup period? SpaceX for example filed with accelerated release of insider/investor shares so I don’t think we can know if this is the case until the filing documents become public.
Look at SpaceXs filing. There is one but it is super short. I was just pointing out that 365day lockup is likely incorrect and OP doesn’t really know that until the filing is approved and becomes public.
Going to give the benefit of the doubt here. I know what lockup period means.
365day lockup isn’t a universal standard. For example for SpaceX 20% of insider shares can be sold in the first few days. 100% within the first 3 months.
Without a public S-1 filing we don’t know what the lockup for Anthropic will be
I'm sure they can get private loads or similiar way to "hedge" those? also dark markets and other tricks exist. Fin. eng. level goes way higher for them, just contact inv. banker or their lower class friends. They will find a way.
Well, you are learning something, just the thing you’re learning has an even shorter usable lifespan than programming languages, namely you’re learning what works to get useful responses from ai agents. Whether or not that has value to you is a different matter, but it’s worth bearing in mind something is being learned, even if it’s not engineering or programming.
> namely you’re learning what works to get useful responses from ai agents.
Having worked a lot with AI agents, I don't agree.
AI agents are amazing at producing response and results that look correct as long as you don't look too closely.
Even when I try to write extremely detailed specs and test harnesses, even Opus 4.8 and GPT-5.5 on max will find creative new ways to write code that breaks under real use cases.
Doing throwaway LLM output, playing with it a little bit, and then calling it done will create a false sense that you're really good at getting LLMs to produce working things.
> You're learning to manage idiot savants, which is a very useful skill.
I think the real bifurcation is whether you will settle on that belief.
Some of us are settling on the belief that the idiot savant, lacking the coherence of a functional mind, cannot be managed. It's essentially a chaos agent masquerading as something more cooperative.
> Well, AI costs are definitely going to go down at least 90% in the next ~18 months for the same quality of output (and probably 90% again in the 24 months after
As far as I can see, token costs have been steadily increasing over the past few months, so I’m not sure that buying the hype that another 90% cost reduction is just around the corner is warranted.
Doesn’t seem like token costs, specifically, are increasing.
Opus cut its token pricing by 66% 6 months ago and it had previously been that higher price consistently for a year and a half (since that model launch).
GPT’s latest model is harder to track since it’s not named, but it’s historically inline with its history.
Not to mention what’s happening with other models like DeepSeek, GLM, and Kimi.
It seems to me the bigger change in costs is based on token appetite. People are discovering agentic capabilities are stronger than they used to be and use cases have broadened because of that. They’ll eventually discover too that these alternative models offer 95% of the intelligence at 20% of the price.
Based on some napkin math, that would be about ~100 watt hours of electricity on an H100 cluster, or, roughly the same amount of energy needed to boil a kettle for a cup of tea.
I think one can do better with a wavelet, shearlet, or curvelet transform that is adapted to the problem domain at hand. But the uncertainty principle still haunts those transforms, and anyway the goal is to be domain-agile.
reply