binary132

What LLM / image / video models would this allow people to self-host that are worth it at the presumably somewhat low speed?

I still seek the fount of infinite wisdom.

Productivity isn’t the only reason to be smarter. For example, there’s also the benefit of asking fewer inane questions on the internet (hypothetically)

could you explain how to optimize prompt caching or point to a doc about it?

I think it has a lot more to do with the Bun situation.

I think it’s mainly a question of whether the price-fixing of VRAM continues or whether an inflection point is forced by the low margins of the industry and potential supply increases. Once the normal scaling of hardware and prices resumes, it’s game over for proprietary, which is why there’s so much urgency to seek market control instead right now.

It can be both — I’m sure the administration would like our stock market not to get ruined by foreign undermining of American AI research companies, and I’m equally sure Anthropic would like to achieve regulatory capture.

maybe it could be running a much more powerful system than you have access to

I don’t think the timing of Trump’s recent sabre-rattling at China is any coincidence

oh dear, are we doing repeating the marketing copy of large companies uncritically

It’s a nationally-sponsored arms race with Western proprietary AI, so as long as the capital sponsors of Western AI lose their shirts and Chinese LLMs become ubiquitous, I think that qualifies as a win condition. Individual labs are no different than running a local restaurant — yeah, it would be great if you land a massive success story, but just making a bunch of money before moving on to the next thing is cool too, even if it’s just via state-sponsored venture capital.

If all or most LLMs are susceptible to failing this test, why would an LLM be a good means of evaluating performance on this test without being given a rubric?

Concepts based on this take no account whatsoever of the differences in currencies and local/national economies that lead to incredibly different costs of normal life in different towns within a state, let alone between different countries around the world. If it were somehow possible to magically level USD incomes of all families depicted in this chart, perhaps to an arbitrary median of $2500/mo, many of the poorest families would be very wealthy — or at least very well-off — by their local standards, while many of the high-income families would lose their homes, healthcare, and means of income due to their local costs of living. It’s always so strange to me that people seem unable to comprehend this simple fact that nearly all of our international economics today are predicated on exploiting.

and in a world with proprietary world-dominatingly powerful AI, who do you think would be the ones managing that? “the government” would just be a thin proxy layer for whatever the magic box decides

1. Tech abuses its users on purpose and everyone knows it.

2. Tech companies are built on eating someone else’s lunch, mostly, and it’s not usually a fair redistribution of value.

He described a separate machine where the execution and context would be hosted, not local sandboxes. Did I misunderstand how Gondolin works?

Uh, they are doing that because they think one of them is going to build the monopoly that will rule the rest of the future approximately forever.

it has to be bait

please let it be bait

it’s almost as if they are just lying about the purpose and doing it for some other purpose that it is perfectly effective for

Don’t forget getting targeted for scams by slopbots on forums and chat platforms

It’s kinda funny even though it’s probably slop-augmented because at multiple points throughout the narrative I found myself second-guessing my belief that it was satire.

Perhaps it was a rhetorical question.

the headline literally explains that it is to maximize hypothetical valuation

The bigger issue is the screening filters are flooded now (and also largely AI “enhanced”) so getting real signal through the noise is becoming basically impossible.

I suspect people who think they are getting away with this are far more obvious than they realize.

This is exactly what first came to my mind as well. Pretty funny scenario to imagine

Do you really think free computing will be banned to the extent that most people cannot easily obtain open models? That seems like a way huger problem than openAI going under

All 550K diesels with modded emissions are not rolling coal, that is a wild extrapolation

I didn’t ask the multiplier of badness of a single individual doing a bad and stinky thing, I asked what you think the _scale_ is. Do you believe that all people with trucks modified to do this are doing it at all times? Or even half the time? How many people do you think are doing it?

that seems like possibly the most unlikely outcome

do you really think there’s no way to prevent or penalize that behavior without preventing the user from owning and operating their own engine?

also, what scale of harm do you think exists from those people?

do you really believe that control of one’s own engine should be removed from all vehicle owners if a few people misuse it?

do you understand that vehicle manufacturers use their proprietary systems that control the vehicle to exploit customers?