Browse Comments

I probably have similar usage patterns as you do (Claude, Codex, etc.), and yes, agents do the darndest things, but I disagree with your characterization. Any system like this will have a percentage distribution where it does what we want, somewhat does we want, or completely ignores us, hallucinates, etc. These percentages have been continuously shifting towards more useful tasks being achieva…

So that’s the line. AI generating images of children. Not the fact that Trump is known to have sexually assaulted a woman and is accused of many more and is a well known friend of Epstein and made comments about young girls. That’s not too far?

We definitely need an alternative that isn’t motivated by profit. AI could be really good, it could be really bad. I don’t trust leaving it up to the Big Tech companies will make it really good for anyone apart from their shareholders…

They are not. The goal of a conflict is control of a region. If the region is damaged, the total potential value of your victory decreases. If there is a nuclear exchange- there is a real risk for a total loss of habitability on earth - and whatever problem you were trying to resolve before is now eclipsed by the problem of surviving in a nuclear wasteland. So nuclear weapons are not efficient …

I don't use chatgpt frequently enough to have plus or have data to export, but this was excellent information for me as I am now excluding chatgpt from my usage entirely. Thankyou.

This reminds me of a video I saw on YouTube (might have been Midwest Safety or code blue cam) that was about a guy at a casino that was arrested for trespassing because someone that looked like him was trespassed for sleeping in the casino.  Despite providing a DL to prove his identity, they went off their super fancy new AI powered camera system that identified him with a 99% certainty score.  C…

Remeber when grok was set to glaze Elon so hard it was embarrassing, or when Elon said the “the only way to eliminate poverty is with the Optimus robot. “ I don’t know how you can be a billionaire, have the chance to be a trillionaire, with 7 children and still be this shallow, immature and out of touch with reality. But that’s the theme these days it seems

Quality of software is already declining. But the real issue will be in the comming years when maintenance comes along, debugging that weird edge case, weaving in a new feature. I'm a software engineer and use AI a lot, but it has to be used for the right things by the right people. Just generating code and jamming it prod.because it works now is going to introduce so much technical debt, googl…

I work in FAANG, I 100% believe them. If anything 75% is low. We have directives to only use AI coding now. And its tracked. It really fucking sucks tbh. No one knows what's getting shipped anymore.

It doesn’t matter, there are no jobs waiting for them. Their parents will soon be unemployable as well, so they won’t be able to afford the AI girlfriend subscription either. Multiple problems solved.

I am 3 seconds into this video and I have to point out: Eliezer Yudkowsky is just Some Guy. He is a blogger who managed to befriend some of the world's most powerful people and a lot of his predictions are just pure science fiction. He has fallen head over heels for Roko's Basilisk, which is Pascal's Wager for dorks. I think you're right to be a little bit dubious of the book's more fantastical …

Such a helpful video! If you’re interested in AI mentions, AICarma is something I’d definitely recommend for keeping track of your brand!

The whole “Japanese Salt Trick for losing weight” ads drive me crazy. They are clearly AI generated because the lower half of the celebrity’s face looks weird, and voices aren’t quite right. I wish YT would remove those, but it makes them money, so they won’t. 🤬

Not quite, I am currently doing consulting outside india, the businesses are adopting ai rapidly, 90% of my customers have 1 question, how to improve ops with ai, non of them ask how do they turn eveyrhing ai. The Analysis done is mind blowing. Most ops staff aren’t required( customer service, admin, filling ops officer ). Ai amplifies the output of each employee by 12x so now instead of hiring…

"Our data" is right. That's why we need to be demanding an AI Dividend for all!

It's crazy, we're using AI in the hopes that it will lead us to working a bit less, or not working so hard, both physically and mentally. But it could get to a point where AI does everything for us, so we're basically useless. Then maybe atrophy kicks in, and we grow physically and mentally weaker, which makes it that much easier for AI to defeat (or enslave) us.

My problem with the "philosophers will figure out whether this qualifies as 'reasoning'" line is that it still assumes way too much about what the models are actually doing, that "oh, yeah, maybe it *technically* might not be reasoning but hey, it's definitely in that direction". No, it's also possible that LLM chain of thought-style reasoning is ultimately a dead end, that it can sort of do thin…

Dissabled artist here, often times I have to stop drawing for periods because my hands shake to much. Does that stop me from drawing as soon as I can? No- Do I turn to Ai? No- I take care of my body so I can make something I love. It’s hard but we can overcome challenges. I feel sorry for people who use this excuse, because if it’s from a dissabled person they probably gave up. And I feel sorry…

My team works on Tammy AIand we have already come to the realization that AI must be controlled or we risk "summoning the demons" in the words of Elon Musk. We definitely need more regulations in this space.

Another note on Firefly - They've also found Midjourney in there as well. So these sources are cross feeding each other as well. Adobe may never have intended this, of course, but if you're using user submitted data for your model, and there's another model that happens to be able to shit out a ton of content, well.....