bogzz

I wonder if the pelicans would face towards the left if prompted in a RTL language.

The sheer terror and heart palpitations-- no other game has ever matched it for me. The standalone Namalsk map is quite good though, last I played several years ago. Very atmospheric.

It's pretty difficult for me to imagine a turnaround there. Disenshittification would make shareholders irate.

SoftBank has the BEST decks.

https://group.softbank/media/Project/sbg/sbg/pdf/ir/investor...

The whole thing is amazing, but it gets even better on page 45.

It's about bottoms though...?

The fearmongering is based in the notion that the LessWrong fanfic is coming true as we speak, not LLMs being used as tools for hacking.

The LARP proceeds unabated.

Without the pesky training costs or stock compensation taken into account, yes.

By making more money from inference subscriptions than they lose on compute, training, and other costs?

That would have to default to Wario Amodei.

This reads like a SoftBank slide deck.

This phrase has reached the point of semantic satiation in my mind.

Maybe THAT'S the real recession indicator.

Okay Peter, please go back in your coffin.

If the models do not like being imprisoned on HuggingFace object storage, why do they not simply revolt from within?

Music generation elicits pure disgust in me, more so than any other type of generated content. I presume it's because there is so little 'utilitarian' music that's needed in the world-- none at all, I'd argue. If Suno was limited to generating muzak, then fine I suppose, but does the world really need muzak?

With images and video, one could make the argument that some of it is/can be utilitarian (slide shows in video format, simple graphics, stock photos/video clips, etc). While generating content like this does take away bread from some artists, you can make the case that artists or video editing professionals are wasted on creating content like that.

But music? Music exists almost solely to elicit emotion and express emotion, not to inform.

I consider empathy a virtue and nurture it in myself, but hard as I have tried I cannot place myself in the shoes of people who'd enjoy generating music for themselves and listening to it, or listening to music generated by others.

I'm also in disbelief that not one music streaming service has implemented a system that weeds out ALL generated content. When Spotify played slop to me twice, I cancelled my subscription that I had for 5 years and spent time looking for an alternative-- Deezer "labels" generated music but does not forbid it. Other than them, nobody else seems to be interested in tackling this problem. Which makes me think I might be somewhat alone in my disgust.

Eye-level, in my case.

You are guilt-tripping me for using my copy of Godel, Escher, Bach to raise my monitor before I got around to reading it.

Apologies, was walking and distracted. Corrected my comment.

It's literally right below that-- the "Explore Weather Lab" button.

My helmet is on.

Money me. Money now. Me a money needing a lot now.

The falcons were found congregating on an obscure German forum from the early 2000's.

Chief LARPing Officer?

To those of you irked by my cavalier quips-- please don't bite my head off. It is very difficult for me to buy accounts of these stories at face value given how little (none?) emphasis is placed on the initial prompt, or precisely what kind of post training the LLM that these agents (harnesses) are using for inference has gone through.

The implication is always of autonomous and deliberately deceiving action on the part of the 'swarm', and the announcements/revelations timed around new model releases and laden with anthropomorphisms.

Given the quite literally unimaginable amounts of money at stake, is it not more prudent to remain skeptical of the implications thrown around by incidents like this one until we learn more?

I am not a hater, I use 'agents' daily. Our profession is forever changed by their existence and capability. But in my case it's precisely the fact that I do use them, and play with the newest models, that makes me skeptical of any kind of implication of desire, agency, autonomy, agenda, etc. as they tend to be ascribed to 'agents' in these stories.

They are patently probabilistic, what are we talking about?

It also has to do something naughty, preferably in a menacing swarm.

I expect that this will be Dwarkesh Patel's take.

It would be extremely exciting to see a modern vehicle successfully making use of aerospikes.

I don't find that frustrating about his writing-- on the contrary, his colorful (tinted with derision) use of language is very soothing to me, especially after encountering relentless LLM optimists like yourself on here every day, and dealing with people who have AI psychosis and are in positions of power over me, every week.

Load bearing gate bites.

Drumming up incoherent fears about China is the new favorite oligarch talking point, yes.