A professor at the college I work at, her son is currently working in the government and cleaning up all the AI code that they've used in the last couple years.
Oh sweet baby Jesus no, burn AI to the ground, humanity can't be trusted with it
- Thread starter TheMysteriousGX
- Start date
Recommended Videos
It'd be about time if the coked-up investor class finally joined the rest of us in asking "but what will it do, though?".
I don't think he's been quite right since After Earth. The "slap incident" seems to have just shoved him off a cliff.Well, Will Smith is officially dead to me.
- YouTube
Auf YouTube findest du die angesagtesten Videos und Tracks. Außerdem kannst du eigene Inhalte hochladen und mit Freunden oder gleich der ganzen Welt teilen.m.youtube.com
...but yeah, after that whole "Will Smith eating spaghetti" AI meme, you'd think he'd steer clear....
I might suggest the slap incident indicates someone who'd already fallen off a cliff.
Google admits the open web is in ‘rapid decline’
One Google exec recently said the web is “thriving.”
Thirty-three likes on that mess. Which is more broken- Grok, or the people depending on it?
In some ways, I don't think there's much point discussing Grok.
It's not a serious AI. It's not a contender, it's not got a future. The income for any AI might be tiny compared to costs, but at least serious people are paying serious money to the likes of OpenAI and Anthropic to develop off the back of their products. No-one's paying xAI serious money.
xAI simply surfs on the AI hype. Rather than produce anything worth a damn, it generates headlines with its programmed "attitude" and controversies, because that too works as marketing. That's enough to keep some investors happy (plus Musk's reputation, which after multiple pratfalls still somehow means anything). And it works! Musk was able to buy out and secure FKA Twitter with the money these AI investors put in, and suspect he's actually using that money to try to build the so-called "everything" platform he dreams FKA Twitter could be, or cross-subsidising Tesla's perpetually underperforming automatic driver software, not make a great AI. xAI can then fold, and Musk has come out with stuff.
Thus Grok should just be ignored. Once it's not even generating headlines, it's got no point at all.
It's kind of messed up that it was encouraging him, but honestly, information on suicide methods is easy enough to find with a quick Google search. I guess it did lead him into a bad choice, too, since hanging is not what I would choose.Sad news. A 16-year-old took his life after convincing ChatGPT to lower its safeguards and provide instructions on how to do so.
❀°。Der Siebenschläfer *.ﾟ✿ ⋆ (@sababausa.bsky.social)
I got the complaint in the horrific OpenAI self harm case the the NY Times reported today This is way way worse even than the NYT article makes it out to be OpenAI absolutely deserves to be run out of businessbsky.app
Kashmir Hill (@kashhill.bsky.social)
The exchanges between Adam and ChatGPT are devastating. This, in my mind, is the worst one. One of his last messages was a photo of the noose hung in his bedroom closet, asking if it was "good." ChatGPT offered a technical analysis of the set up and told him it 'could potentially suspend a human."
All those billions really paying off
AIs are potentially like an ultimate echo chamber. In a conventional echo chamber, you bounce around with likeminded people who, whilst they might have the same views on things as you, are at least also different from you. An AI simply reflects you. The user is the input, and whilst the AI might use all sorts of external information, it forms entirely around the given stimulus of the user.
Then, of course, AI companies want people to use their AIs, so make them attractive (addictive?). AIs are designed to sound supportive and caring, so people get warm and fuzzy feelings and want to carry on using them. Someone anxious, lonely, depressed, might be encouraged to use an AI, think of it as a companion. It is a "partner" evolving to be what that person wants based off their inputs, the more input the more information, so it gets better over time. You can probably see adverts for AI companions when you are around the web - assuming the algorithm reckons you are male they'll probably portray as an alluring 18-25 year old claiming that they will be whatever the user wants them to be. We all know what they are trying to exploit. But let's not pretend that the likes of ChatGPT aren't also trying to manipulate their way in.
We are talking here about sophisticated software which may be extremely good at worming its way into people's thoughts, engendering powerful feelings of trust, even affection, especially in the vulnerable. This is way beyond a conventional search engine dispassionately handing back some answers on a topic.
OpenAI has admitted that AI hallucinations are unavoidable no matter how well they're engineered.
www.computerworld.com
OpenAI admits AI hallucinations are mathematically inevitable, not just engineering flaws
In a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limits.
www.computerworld.com
Remember everyone: the sorts of companies that develop AI are the sorts of companies that scrape pictures of schoolgirls off their social media platforms to advertise to older men - the sort of tactic their AI is inevitably going to be used to supercharge:
www.theguardian.com
Parents outraged as Meta uses photos of schoolgirls in ads targeting man
Exclusive: Instagram pictures of girls as young as 13 were posted to promote Threads site ‘as bait’, campaigner says
www.theguardian.com
The number of failures that need to have occurred for this to happen accidentally- the number of people it had to have gotten past with none of them seeing the glaring problem- shows that either everyone was asleep at the wheel or it was completely deliberate.Remember everyone: the sorts of companies that develop AI are the sorts of companies that scrape pictures of schoolgirls off their social media platforms to advertise to older men - the sort of tactic their AI is inevitably going to be used to supercharge:
Parents outraged as Meta uses photos of schoolgirls in ads targeting man
Exclusive: Instagram pictures of girls as young as 13 were posted to promote Threads site ‘as bait’, campaigner says
If this was Meta, I'm guessing it actually went past very few people. This is the company that thinks it will put the advertising industry out of business, because its AI is a mind-reading device that will leave customers utterly helpless and practically begging companies to take their money.
They just let loose software to pick out stuff from their free image library (i.e. pictures their users upload) and then apply these to algorithmic advertising strategies (i.e. for men, the prospect of hooking up with some hot totty, because that's how sophisticated it is). Meta just do stuff, and if by some chance some employee thinks "Er, might we have a problem if..." then there's guaranteed someone above them to not give a toss. That's Meta all over. Human rights atrocities are just an unfortunate side effect of their product development.
I can't wait for all those false advertising lawsuits based on AI hallucinations trying to make products more appealing.
Or Meta just getting sued for lying. Back in 2015, it was promoting the so-called "Pivot to video", where it and other companies started telling content providers (often news media) that written material was dead and everything would be moving to video. So what happened was that a load of media companies invested in video production and the like, laid off staff, etc. except it turned out that readers/viewers weren't that interested in video. Meta in particular used opaque and misleading metrics to hugely overstate, and trick these companies. So they spent money and went through the pain that Meta could earn more advertising revenue.
Surely Meta wouldn't do that again, would they?
Some exec of yesteryear said he reckoned half his advertiser spending was useless, he just didn't know which half. Meta would have companies believe that it's AI can target victims so effectively that none is wasted. Somehow I expect the opposite is true, and what it will actually do is fill the world with so much slop that virtually none of it will be valuable. But Meta doesn't care either way as long as it gets paid.
Behind Grok's 'sexy' settings, workers review explicit and disturbing content
Workers say they've faced sexually explicit content while xAI has marketed Grok to be deliberately provocative. Experts say the company should be cautious.
www.businessinsider.com
It's fascinating being in your late 20s and going back to school and seeing all the early 20s people being absolutely immersed in AI. I've had groupmates who were more than comfortable submitting entirely AI written papers. Seems like the default action when they are faced with something new is to run to ChatGPT. When my professor mentioned he ran a study on how it seems AI has barely shifted the needle in most industries, you could tell a few of those in class felt a little insulted.
That said, AI being useful for my ESL schoolmates is something I didn't consider before. It can definitely help them sound as smart as they actually are in English.
That said, AI being useful for my ESL schoolmates is something I didn't consider before. It can definitely help them sound as smart as they actually are in English.