Artificial intelligence still has some way to go

so glad these robots can have this discussion amongst themselves

I'm the real deal Brian

it seems that everyone hates this stuff and yet we march forward

new civ description

"Thank you for understanding my mental health struggles"

https://www.theguardian.com/artanddesign/2023/oct/02/ai-vincent-van-gogh-talks-of-mental-health-struggles-in-paris-exhibition

jesus christ

was just reading a good article about that, arguing that AI is already a net negative and will likely just get worse

https://charliemoe.substack.com/p/at-least-the-robber-barons-built

the first paragraph:

Recently, I came across this article about how searching on Google “can you melt eggs” returns an answer of yes. This was due to the fact that the website Quora was using an AI chatbot to automatically reply to some more basic inquiries from its users, and even when that answer was wrong, it still got promoted to the top of Google’s search results. A more gruesome example is that MSN recently had to pull an AI generated headline that announced the death of former NBA player Brandon Hunter by declaring him “useless at 42”.

ties into what treeship says, everyone hates this shit, but it's still taking over the internet and it doesn't seem like it can be stopped. pretty grim.

it does amaze me the kind of visceral reactions people get when they do things like AI-generated Calvin and Hobbes strips. they're not good, but they look kind of legit. nobody's trying to sell them or make them as anything more than a thought experiment. regardless people just fucking hate them. I've never really seen that sort of reaction to a piece of technology before. even the Metaverse shit was at least kinda funny.

I have a theory that the angriest people are heavily invested in making and/or consuming fan art, and it pisses them off because it reveals how cheap and unimaginative it all is

“Cofounder & CEO @ProjectAI”

Late stage empires don’t build things this extraordinary.

pic.twitter.com/H60w5vGKHi

— Axial Wanderer (@EricWollberg) September 30, 2023

*@PropheticAI, which is even worse

I think AI finally no longer has much of a ways to go:

https://www.instagram.com/p/CyNISqMrnM4/?img_index=1

I refuse to be impressed until AI creates wholly satisfying pics of dogs playing poker. those i've seen fall far short of the magisterial qualities of the original.

AI song generation still has some way to go but this is pretty cool (one riff at a time)

https://www.riffusion.com

i expect ai to be capable of superhuman persuasion well before it is superhuman at general intelligence, which may lead to some very strange outcomes

— Sam Altman (@sama) October 25, 2023

lmao

dilpilled

I actually got a demo of an AI tool for technical proposal writing the other day. Started out skeptical. Ended up, weirdly, impressed. It was pretty damn good at doing a first draft.

Of course I had to lie down with a cold washcloth on my forehead for an hour afterwards, watching my career and livelihood slip away in a few minutes.

Once I recovered slightly, I started coming to terms with the idea that the rest of my career will necessarily involve editing and/or fact-checking AI-generated text.

Like it or not, it's here to stay. So I guess I need to deal with it and make a slightly different case for how I can add value, stay employed, feed my children, etc.

Somebody at work made a an audio deepfake of me speaking Chinese and you know, it sounds pretty fuckin good

Not only did @Gannett's AI posts sound robotic - but their “authors” may not even be real.

They’re not Reviewed editors, don’t have LinkedIn profiles, and Google searches turn up no other written work. pic.twitter.com/XkFGhqwTuO

— NewsGuild of New York (@nyguild) October 24, 2023

I liked this CS50 talk by AI startup dude/former Harvard lecturer Matt Welsh. He's pretty honest about how effective LLM's are as job killers lol

https://www.youtube.com/watch?v=JhCl-GeT4jw

It seems like a pretty good prediction of the way things are going. The big question in my mind is whether this is going to happen in 5 years or 50.

My bf works in graphic design, and he was saying how disconcerting it is now that the latest Photoshop has this really sophisticated AI tools where you can just like, replace someone's face with someone elses without having to know how to lasso/feather/do the layers. It does it all for you and does it very well.

But I guess if you were going to do that process anyway, and it just does it quicker, is it really that a problem?

"IS IT REALLY THAT A PROBLEM". ffs alcohol give me a break

The best thing about the new AI built in Photoshop is now I can use one photograph and make it work in any aspect ratio in an ad campaign.

https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72cb6982-81c3-4e58-ba91-ef0e1cacbe3c_1170x1786.jpeg

Ah, fuck, anyway, supposedly driverless cars needing secret remote human intervention every couple of miles:
https://garymarcus.substack.com/p/could-cruise-be-the-theranos-of-ai

“Those vehicles were supported by a vast operations staff, with 1.5 workers per vehicle”

haa they need more drivers than a regular car

lololol perfect

yep my brother-in-law is also in a specific type of graphic design, doing really well, successfully built up his business working for himself. and will probably be out of work in a couple of years.

I expect a lot of creative workers will shift from creating-from-scratch to something more like curating. We'll be selecting, tailoring, customizing and editing AI-generated content.

I talk about this with my boss (still a human, a far as I can tell) a fair amount. In a way it's similar to what I do when managing a team of human writers - I still need to review what they produce, select and tailor it to the situation.

The thing that sucks (for me personally) is that the steps that can be automated are the fun ones (actually creating). That leaves the un-fun stuff like managing. I still have to do that part.

I'd prefer it the other way around. I used to ask people: Don't hire me an assistant, hire me a boss.

hmm, good point. I think that outside of staffing at my workplace, the big problem right now is determining project priorities and assigning work because there’s too much to do and no one’s setting priorities. maybe I should just dump all of the project backlog into a LLM and ask it what to do

https://cdn1.suno.ai/5e30c14b-dea8-4b4d-aefd-2767e3305574_0.mp4

"a yacht rock song about how people from different generations listen to indie rock"

ai is funny it makes things that are impressive for being able to produce them with a one line prompt but are in quality so so bad

presumably most of the music that that program was trained on was better than that, chat gpt consistently churns out the most didnt do the reading bs prose, the image ones all have this corny dramatic lighting and no sense of composition, when you average everything out with no sense of taste what you get is worse than all the source material, funny to think about

The corny dramatic lighting seems to be beloved by AI image enthusiasts, and perhaps that drives the developers of the algorithms even if they don't love it themselves (which they probably do)

or maybe it just spits that out for reasons due to the nature of the technology idk when it was just one program doing it i figured someone had made the choice but there are more of them now and theyre all like that

https://www.cnbc.com/2023/11/17/sam-altman-leaves-openai-mira-murati-appointed-interim-boss.html

wow

Has chat gpt commented on this yet?

it denied it. then said sorry for mistake, it did indeed happen.

he cried like turtle

huh. sam altman forced to step down, unexpectedly

"Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities," the blog reads, in part. "The board no longer has confidence in his ability to continue leading OpenAI."

https://arstechnica.com/ai/2023/11/openai-fires-ceo-sam-altman-citing-less-than-candid-communications/

very interesting!

heh

Anil Dash

How is the OpenAI board gonna fire Sam Altman just because he gave answers that sounded plausible but weren't actually accurate?

Light accounting fraud or doing stuff contrary to the pseudo philanthropic mission of the board. Impossible to say.

they seem pretty mad

we can't rule out a tense meeting in a dark room with a large bright screen where sam altman declared that he was going to launch his physical ai war by taking over the agricultural sector

[q]https://www.cnbc.com/2023/11/17/sam-altman-leaves-openai-mira-murati-appointed-interim-boss.html

wow

― 龜, Friday, November 17, 2023 2:42 PM (one hour ago) [/i]

and shoot, sorry 龜 for missing this!

they found out he was drinking all the water that was supposed to cool their servers

...OpenAI’s board includes chief scientist Ilya Sutskever and independent directors like Quora CEO Adam D’Angelo

very interesting that ned exposes quora christian infant viral internet to bluesky and suddenly sam altman steps down

would you ask a christain baby to do your homework for you

google man riding hard for alt man

Sam Altman is a hero of mine. He built a company from nothing to $90 Billion in value, and changed our collective world forever. I can't wait to see what he does next. I, and billions of people, will benefit from his future work- it's going to be simply incredible. Thank you…

— Eric Schmidt (@ericschmidt) November 17, 2023