Gotta find the right hostels

Yawn.

I have the best model, but it lives in Canada, so you wouldn't know it.

This. We use both opus 5.5 and 6 Sol in a way that allows us to compare their performance. Opus is considerably better.

Ah shit, here we go again.

You mean what they said about GPT-2?

https://www.youtube.com/watch?v=DbOckHiVXZc

I have a project that requires a model with at least a 20% chance of existential threat. Please release this Sama.

Because no car company would ever announce they had an unsafe car. They’d fix the problem and announce an awesome car.

Hype: you tell the world you alone have a very powerful and potentially dangerous product, and you're the only one who can be trusted keeping it safe.

I would consider it more "bizarre" to assume that Sam Altman was telling the truth about anything.

https://web.archive.org/web/20240112010809/https://www.sfchr...

There are cars that market based on only being track legal — and even more that market on requiring being tuned down for street safety.

Most sports cars do the second, exactly like OpenAI.

At the moment, their main "customer" is investors, not users. But even users want to use the most dangerous model, because dangerous is just another word for powerful.

Sports cars are advertised with how fast they can accelerate from time to time, or tesla's "ludicrous mode" - that is them advertising how dangerous it is.

Car companies sell mostly the same car every year. They don’t have billions (or trillions) of dollars of debt riding on the next model they release being substantially better and cheaper (closer to profitability) than the previous one.

These AI companies have been selling AGI as coming any day for a while now. If it doesn’t arrive soon, the safety angle may be the only spin that keeps the massive (and required) investments pouring in.

If instead the narrative became that LLM progress was slowing, we’d almost certainly be looking at the next global recession.

At this point, there is too much money in AI for the truth to have much of a chance.

I remember when they promised to slow down and released a new model less than a week later.

Isn’t this like the twentieth model they said this about?

Are they sure it isn’t because they are setting money on fire and have no business model?

My lobster is too buttery and delicious as well.

Earlier source: https://news.ycombinator.com/item?id=49885133

OpenAI will have to keep releasing unless it wants to bleed customers. If Anthropic leaps ahead by a couple of more models, even more OpenAI users will then switch over their subscription from OpenAI to Anthropic.

The safety concerns exist only because the underlying third-party servers are grossly insecure to begin with.

Really inspiring comments guys, really looking connected to reality on this one.