Demigirlboss

@demigirlboss.win

Lover of FOSS, hater of proprietary standards, ATProto's nineteenth biggest glazer, and unironically thinks RISC architecture is gonna change everything. Described by critics as "a BAD person". Wrangling hungry ghosts in jars at @fieldleveltech.org.

Some people seem to think that ML wasn't originally considered AI and are criticizing labeling ML as AI as something recent. That's not right. ML is a subfield of AI. Not all AI is ML, but ML is AI. The history on this relationship is clear; they've always been associated this way.

we don't have time for another plague apocalypse. we already did that. we're on to robot apocalypse now, try to catch up. next in line is fungal apocalypse. please consult the schedule

DVI port prion disease: a Google tale a DVI plug had one pin bent when inserted into an outlet, it bent one of the plastic sleeves the next DVI plug in that outlet bent in the exact same way soon: most DVI ports in a building or two didn't work anymore

Apple Music always assuming that what I want right after a Hozier album is Ed Sheeran is some confidence I wish I had

I was talking to @esteraxiom.mom about how voice training enables you to expand your repertoire of bits significantly and she said, "You can do a sexist comedy duo all by yourself!!!!" So naturally I went, "I HATE MY BITCH WIFE!/I'M ADDICTED TO OPIATES!"

creating regulations to regulate the impossible like "emotional reliance" or "unreasonable decisions" is how corruption and oppression happens similar to US's badly defined obscenity law where you can't know you've broken it until you are charged for something "a reasonable" person finds immoral

Hetan Shah@hetanshah.bsky.social · 15h ago

Interesting to see how China is regulating AI companions including saying that chatbots cannot create ‘emotional reliance or addiction’. Fines are relatively small but seems like some companies are taking it seriously www.bbc.co.uk/news/article...

Chatbots can't create "emotional reliance or addiction" or induce people to make
"unreasonable decisions". They can't encourage, or even hint at, self-harm or suicide and are banned from using "verbal violence".
Companies can face fines of 200,000 yuan
(E22,000) or the risk that the government
"circulates criticism", in effect publicly shaming them.
A week before the law came into effect, two of China's biggest tech companies, Alibaba and ByteDance, shut down the companion features on their Al assistants.

Something not sufficiently registered in the convo around AI/IP: the value that was gained from pretraining (on our collective legacy) is now common property. People do pay more for frontier models than open ones — but the differences they're paying for, at this point, come from post-training.

Ted Underwood@tedunderwood.com · 14h ago

Fortunately, that bridge itself is easy as heck to copy and make free. At this point pretty much everything that was gained from the initial pre-training is in fact, universally shared. Differences between models that people pay for are now all post training.