The timing and similarity highly suggests this is a problem with how almost all software has implemented the webp standard in its image processing software.

Did you read the article or the post? The point was that both places where the vulnerability was found probably used libwepb. So it’s not that there’s something inherently vulnerable in handling webp, just that they both used the same library which had a vulnerability. (Presumably the article was a little vague about the Apple side because the source wasn’t open/available.)

given that the programs processing images often have escalated privileges.

What? That sounds like a really strange thing to say. I guess one could argue it’s technically true because browsers can be considered “a program that processes images” and a browser component can end up in stuff with escalated privileges. That’s kind of a special case though and in general there’s no reason for the vast majority of programs that process images to have special privileges.

It’s a briefcase full of cash.

I’m pretty sure you could just say “It’s tax free” or even double the amount to $2 million and it wouldn’t really change which people would do it and which wouldn’t.

I’d do it, as long as I was really convinced that the only danger was mental, not physical.

You can wing it with baking, at least for some types of stuff. Oatmeal raisin cookies don’t really take precision, as an example.

This is a really misleading title if it’s just grouping places where people were imprisoned with places people were actually tortured. There’s obviously a massive difference. This seems like the original article in Ukrainian: https://mvs.gov.ua/news/pid-cas-zustrici-iz-specialnoiu-dopovidackoiu-oon-z-pitan-tortur-katerina-pavlicenko-povidomila-pro-viiavlennia-v-ukrayini-80-rosiiskix-kativen

Are they actually saying people were definitely tortured in all 80 places there? (Also kind of funny, Google Translate seems to do a better job than the link in OP but it’s still not clear to me exactly what they meant.)

I think you might have misunderstood the post? There wasn’t anything about the manufacturer disabling the radio. The person I replied to said they’d choose not to use the car’s fancy features and I thought it was funny they’d do that to “spite” the manufacturer after giving them a whole bunch of money.

Maybe. I feel like it’s going to be kind of hard to make them care if you’re still buying their product though.

I’ll plug in my phone, ignore your entire. Infotainment and actively campaign for it to fail and blow up in your face.

This sounds kind of funny. “I’ll spend $60,000 on your car but I won’t turn on the radio. That’ll show you!”

The title makes it sound like Rotten Tomatoes deliberately did something shady. What actually seems to have happened is:

1. Rotten Tomatoes aggregates critic reviews. As far as I know, those critics aren’t really affiliated with Rotten Tomatoes.
2. Some of the critics that make up that aggregated rating got bribed to increase their evaluation of the movie.
3. Consequently the score on sites that aggregate reviews like Rotten Tomatoes increased.

As for me, it would have to be a number of people that, by their sudden absence, manifestly affects the life of people I do know and care about. Like, at least a billion or so if randomly chosen

If they ask someone else, you’d better hope that other person doesn’t think like you - for your sake and the sake of people you care about.

I hope you’re just looking for interesting responses rather than a definite answer!

I genuinely wonder if saving a negative number of people would be better overall. Humans, especially ones in developed countries like those privileged enough to be posting about stuff like this are responsible for a lot of negative effects we don’t really like to think about. We benefit from exploiting other people, animals, using resources in unsustainable ways.

I think even if someone takes a lot of individual steps like going vegan, trying to recycle, minimizing transportation and other consumption, not having children, etc that they’re still not even going to break even with the harmful effects just existing causes.

If it wasn’t for effects like that I’d probably say 2-3 but in reality I’m not really sure if I truly should save anyone. (By the way, you don’t have to worry about me going out and murdering people.)

A lot of people don’t understand the limitations/weaknesses of AI. The carelessness was probably more in not actually learning about the tool he was relying on (and just assuming it was reliable information).

So is the question how many it would take for me to believe that I should give my life, or how many before I’d actually have the willpower to choose to die?

we aren’t breaking the event horizon threshold as title suggests

It wouldn’t be pop-sci if it didn’t have a misleading clickbait title!

Then my awkward ass is sitting there knowing I need to say “I’m good. How are you?”

You don’t have to say that.

1. “It goes.”
2. “Another day, another doughnut.”
3. “How’s it going?” - people really won’t be surprised if you just don’t answer at all.
4. “I’ve been worse.”
5. “If I complained, who’d listen?”
6. “Hi.”
7. “Hey.”
8. “No news is good news.”

In the context of random people who don’t know each other, it basically just means “I acknowledge your existence”. Acknowledge their existence and you’re good.

They’ll fine you, and if you don’t pay they’ll put a lien on your house. You can’t really just ignore them.

There are basically two main possibilities:

1. They’re unreasonable.
2. You’re unreasonable.

If it’s the first one, it doesn’t really matter how you respond. The best policy is to avoid dealing with people like that as much as possible.

If it’s the second one then you should work on trying to fix it. That’s the best way to respond.

I’ll give CenturyLink a chance because I really hate Comcast.

I can definitely believe that Comcast can manage to be more horrible than CenturyLink. Not an easy task though.

The next best thing is satellite at $100 for 30Mps, versus 200Mps for $40 for CenturyLink, or Comcast.

Don’t go with satellite if you care about latency at all. (From what I know, it also tends to have pretty pitiful upload speeds.)

I looked at Centurylink’s “compatible” modems and they’re all rated 3.5 stars and cost $150+ or more.

I had their service for 6-7 years and I have to admit I never had an issue with the modem (and the service was generally reliable with little downtime). Their customer service though…

CenturyLink is absolute garbage. I rented a DSL modem from them. It got fried by lightning so they had to replace it. They sent me a modem that wasn’t compatible with my service. A couple years later, I had another one get zapped. I double checked with not one but two customer service reps to make sure they were sending me a modem that worked with my service. They sent me one that wasn’t compatible with my service. Then they took a few weeks to send me one that actually was compatible. When it got here, it either didn’t work or something else in the wiring was messed up (probably more likely).

That last part might not have been their fault but I could have known about it 3 weeks sooner. At that point I didn’t have much confidence they’d get it fixed while I still have my youth and good looks. Fortunately a smaller fiber company had just started serving the area and I was able to immediate cancel the CenturyLink service. More than 3 times faster and slightly cheaper as well. Also symmetric upload is pretty nice. CenturyLink is in for a rude awakening as competition appears in places where they previously were the only choice.

The problem is not really the LLM itself - it’s how some people are trying to use it.

This I can definitely agree with.

ChatGPT cannot discern between instructions from the developer and those from the user

I don’t know about ChatGPT, but this problem probably isn’t really that hard to deal with. You might already know text gets encoded to token ids. It’s also possible to have special token ids like start of text, end of text, etc. Using those special non-text token ids and appropriate training, instructions can be unambiguously separated from something like text to summarize.

The bad summary gets circulated around to multiple other sites by users and automated scraping, and now there’s a real mess of misinformation out there.

Ehh, people do that themselves pretty well too. The LLM possibly is more susceptible to being tricked but people are more likely to just do bad faith stuff deliberately.

Not really because of this specific problem, but I’m definitely not a fan of auto summaries (and bots that wander the internet auto summarizing stuff no one actually asked them to). I’ve seen plenty of examples where the summary is wrong or misleading without any weird stuff like hidden instructions.

Yeah the whole article has me wondering wtf they are expecting from it in the first place.

They’re expecting that approach will drive clicks. There are a lot of articles like that, exploiting how people don’t really understand LLMs but are also kind of afraid of them. Also a decent way to harvest upvotes.

Just want to be clear, I think it’s silly freaking out about stuff like in the article. I’m not saying people should really trust them. I’m really interested in the technology, but I don’t really use it for anything except messing around personally. It’s basically like asking random people on the internet except 1) it can’t really get updated based on new information and 2) there’s no counterpoint. The second part is really important, because while random people on the internet can say wrong/misleading stuff, in a forum situation there’s a good chance someone will chime in and say “No, that’s wrong because…” while with the LLM you just get its side.