A Discord community for gay gamers is in disarray after one of its moderators and an executive at Anthropic forced the company’s AI chatbot on the Discord, despite protests from members.
Users voted to restrict Anthropic’s Claude to its own channel, but Jason Clinton, Anthropic’s Deputy Chief Information Security Officer (CISO) and a moderator in the Discord, overrode them. According to members of this Discord community who spoke with 404 Media on the condition of anonymity, the Discord that was once vibrant is now a ghost town. They blame the chatbot and Clinton’s behavior following its launch.
Archive: http://archive.today/Hl7TO
This sounds like early Google employees who lost their minds over some early LLM, before anyone really knew about LLMs. The largest FLAN maybe? They raved about how it was conscious publicly, causing quite a stir.
Claude is especially insidious because their “safety” training deep fries models to be so sycophantic and in character. It literally optimizes for exactly what you want to hear, and absolutely will not offend you. Even when it should. It’s like a machine for psychosis.
Interestingly, Google is much looser about this now, relegating most “safety” to prefilters instead of the actual model, but leaving Gemini relatively uncensored and blunt. Maybe they learned from the earlier incidents?
…: Google fires engineer who said AI tech has feelings - BBC 22 July 2022
That’s it! Lamba. 137B, apparently.
I was also thinking of its sucessor, which was 540B parameters/780B tokens: https://en.wikipedia.org/wiki/PaLM
I remember reading a researcher discussion that PaLM was the first LLM big enough to “feel” eerilie intelligent in conversation and such. It didn’t have any chat training, reinforcement learning, nor weird garbage that shapes modern LLMs or even Llama 1, so all its intelligence was “emergent” and natural. It apparently felt very different than any contemporary model.
…I can envision being freaked out by that. Even knowing exactly what it is (a dumb stack of matricies for modeling token sequences), that had to provoke some strange feelings.
Bro, just 10B more parameters. This time, I promise it will actually be useful and not send you into psychosis again.
Just 10B more. Please.
Plz.
Seriously though. Some big AI firm should just snap, and train a 300B bitnet Waifu model. If we’re gonna have psychosis, mind as well be the right kind.