"It's unfair to condemn all AI. There's a lot of really useful Machine Learning that doesn't post the same risks and harms as generative models."
Anybody who is still using "AI" to market their research or product is hoping to profit off the "greater fools" of the AI Bubble. So fuck them. If you don't want to be tarred by the same brush, don't use the term "AI" for your own work
Yup, the term "AI" is utterly toxic now and best avoided.
If there's something genuinely good don't call it "AI".
@FediThing @baldur Yeah, really feel bad for the generative art folks.
@stefan @FediThing Yeah. Absolutely. Interesting work and field. Completely coopted by some of the worst people around.
@baldur @stefan @FediThing
Thankfully haven't seen that many people just lazily copy-n-pasting the Genuary prompts into Midjourney etc like in previous years.
Although there was this one guy who asked ChatGPT to write p5js code for every prompt
https://bsky.app/profile/piterpasma.nl/post/3ldh73h3dhs24
Augmented Idiocy
@FediThing @baldur
I mentally substitute “shit” when I see “AI” and that is much more helpful and fun.
“See our new shit interface on our brand new shit platform!”
@baldur AI has been around a long time, and is sort of the "catch all" for computations that seem to work but are not established to work by computer science and mathematical principles at which point they just become algorithms. It's the same sort of logic that says "alternative medicine" that works is just medicine. Because of this uncertainty and mysticism surrounding it, it is easy to peddle AI to solve any problem.
@baldur I think you're basically right (I wouldn't launch anything with "AI" on it) but I don't know of a good alternative word to describe non-generative (strictly speaking still sort-of-generative I suppose?) cases like AI audio transcription, subtitles etc.
@baldur or we can double down and confuse everyone: https://fed.brid.gy/r/https://bsky.app/profile/did:plc:kn4dz4f67maytvsbcnbry36k/post/3lfpesnbqk22e
@baldur i hate this so much. It is difficult to market solutions that provide insights based on statistics with explanations. Strictly non-generative, explainable machine learning - we wanted to brand this as the useful AI. Now we have to resort to these technical sounding terms again. It is annoying as hell, since most C-level peops are not listening then…
@codingGarden @baldur You really don’t.
Hit //FHWL.se up.
The site is work in progress, but we have some pretty convincing cases, for CGI (the consulting giant) among others.
I’m positively sure you do not have to resort to techno speak.
@baldur yup. ‘machine learning’ was plenty good enough for all the cool OCR/HTR we’re doing. and less pompous and overeagerly assuming, too.
“AI” is now a Trojan horse investment intended to (further?) disenfranchise knowledge workers, creatives, and the working class.
The disenfranchised struggle to maintain solidarity.
That is the strategic objective.
The only people unaware (or uninterested?) in the lack of ROI are the investors.
@baldur gosh if only there was another term that meant Neural Net.