Right and which makes it all the worse that the editor-in-chief of a major tech blog would validate the behavior.
(I hope the grandparent comment didn’t intend to be read as accusing people of doing the wrong thing knowingly, especially since so many AI companies intentionally mislead potential customers.)
The difference between the medical industry and the AI industry is like night and day. Medications are tested by professionals, side effects are documented, and professionals recommend them.
The AI/Wellness industry, by comparison, grabs people that should have been treated by the medical system. AI is the medicine equivalent of a weirdo in an alleyway promising that they’re a doctor, giving you some random pills with ingredients unknown to even them, but that they know for sure has caused people to kill themselves before. And the weirdo’s only goal is to make you feel correct about ingesting that medicine.
Regulation would be great, though. In fact, the product should be pulled until that regulation is in place.
Like a director of a social media company that is allergic to taking responsibility for its actions, for example?
Can you explain you believe it’s a “moral panic” when we identify software that has been caught encouraging suicide and homicide in real life…
Edit: While pushing baseless conspiracies that AI will kill everyone?
As predicted, Mike Masnick is the author. Mike has a conflict of interest when it comes to reporting on platforms’ responsibilities, because he’s on the board for Bluesky… The social media company.
And he’s trying to argue that chatbots are good for mental health actually. Never mind healthcare, he praises chatbots.
Yet chatbots have emerged as first aid for people experiencing mental health issues, providing genuine benefit to those who aren’t in crisis but are not OK either. Heavy-handed legislation risks derailing this breakthrough in support, creating more problems than it solves.
The proof? Self-reports. Including people who use the Replika Girlfriend-bot.
At this point, I consider anything on Mike’s website that’s related to social media to be compromised, and this is yet another example of that disappointing pattern.
The comments in the article are actually pretty good. Like this one.
I love how on Techdirt, when it comes to LLMs, the entire concept of product liability just goes right out the window. If this were a physical object that, ha ha, occasionally convinced people to commit suicide or murder, or spiral off into other delusions, it’d be off the shelves in a heartbeat, no matter how useful some people thought it was, and the manufacturer would be rightly sued into the ground. But according to Techdirt, because it’s software, it is now and forever a permanent and untouchable part of the internet landscape and regulating it is impossible and undesirable.
I’m (cautiously) interested in the concept of built-for-purpose chatbots being used therapeutically, although I expect the providers to fail horribly at not abusing the massive trove of personal data they’ll gain access to. But if a corporation can’t produce a general purpose chatbot that won’t help people kill themselves, they have no intrinsic right to just dump it on the internet and say “it’s not our fault.” If that’s a bet they want to make, then they need to accept that they’re going to take their lumps.
ETA: the comment above ended up causing a Mike freakout. It was written by user TheKilt, who is exceptionally friendly and willing to concede points to Mike. Mike responds by accusing TheKilt of lying, and then proceeds to react to different people in the same thread who are only insulting him. TheKilt tries to fetch Mike’s attention one last time, but Mike keeps ignoring him.
Mike is choosing the lowest hanging fruit and ignoring substantive criticism. It’s embarrassing.
Do you just propose the union just adopt Google’s exact stance on this? Are you willing to accept a weaker one? A much weaker one, perhaps? Where do you draw your lines?
I choose option 3. A union with better policies. I don’t see what’s preventing them from providing them.
I find it a bit offensive that you assume Google employees can only comprehend the simplest language, and it’s coincidentally the language handed to them from on high by Google themselves. (Ah. Dot ML.)
But let’s assume you’re correct, and engage in a little creativity to simplify employee complaints in order to make it have fewer loopholes.
Employees pushing back on the deal are concerned
that it could open the door forGoogle’s technology [could] be used forautonomousweapons andmasssurveillanceof American citizens.
12 fewer words, 4 fewer loopholes (preexisting surveillance, semi autonomous weapons, selective surveillance, foreign mass surveillance).
Like I said, the unionization in this instance is mostly good. There’s plenty of examples in the article that I left out because they’re unobjectionable. It’s just unfortunate seeing the union repeat talking points manufactured by their employer.
Here’s a hot take: If a union does bad things, those things are still bad. Like police unions. Hotter take: mass murder is bad.
Surely the brilliant minds at Google can think just a little bit about the loopholes? No?
Unnecessarily long article (which says “4GB” 33 times, and the complete phrase “4GB AI model” ten times)… Once or twice was all I needed.
But the article author(s) came across a good point. If pushed out to ~15% of Chrome users without consent:
And that’s just for the initial data push. Models need ✨updates!✨
This is mostly good; the bad is that the people unionizing would use the same pro-Military Industrial Complex language as OpenAI to explain their willingness to commit violent.
Employees pushing back on the deal are concerned that it could open the door for Google’s technology to be used for autonomous weapons and mass surveillance of American citizens.
The loopholes in this seemingly Wholesome Keanu Chungus statement are broad and pretty obvious if given a little thought. Employees allow for:
If you dig any deeper into it, most of the linked articles from the Business Insider post point to AI Apocalypse fearmongering from Google and Google subsidiaries. Sigh.
The best option on this list that actually runs Linux looks to be the flx1s, which is only $500 and pretty loaded spec-wise. Shout out to @schwim@piefed.zip for actually clicking all the links.
So… Is Trump still banning states from making their own regulations based on real AI problems like environmental and psychological harm?
They’re just going to regulate the industry based on something that, at best, looks like a staged event?
[Anthropic] should immediately stop misleading people through company documentation that models are “blackmailing” people or, as Anthropic did in its Mythos system card, suggest a model has “broken containment and sent a message” when it A) was instructed to do so and B) did not actually break out of any container.
Relevant comment from Hacker News:
This is just a leverage buyout and it will likely result in the slow death of both parties while there is value extraction for those in control. Think Sears, Toys R US and similar.
The CEO has a very specific deal where he gets paid significant compensation for specific valuations, which this is likely to achieve. That is value extraction at the cost of shareholders who will be on the hook for the leveraged loan and which will likely wipe them all out over time.
FauxLiving, the burden of proof is on you to show us why your AI utopian vision will happen as you predict it. A paper does not guarantee your fantasy.
FauxLiving, I appreciate your guarantees about the future, but can you demonstrate why the for-profit medical and AI industries wouldn’t cut corners if the AI behaved the way you hope it will?
I think you’re right. Despite Google’s latest horrible promise to wreck user choice, auto-updates are possible from third-party app stores in Android right now, and they have been for a long tint. And it’s possible for competing app stores to update the same app, provided the update is signed by the same key as the original.
In F-Droid, for example, you only need to manually click the system level Install button once per app; after that it can update apps without asking. But if you download a more up-to-date version of the package from a different source, you can install it there too (if you approve the installation of the app from any source, and not just the first one you used).