AI is very good at some things, and very bad at others. There are scenarios where it is so much better than humans that it essentially eliminates an entire field of study and also scenarios where it is laughably bad (you can find almost infinite examples online).

All LLMs are AI, not all AIs are LLMs. Do not confuse the two.

Since you are obviously an AI sceptic, I would assume you avoid LLMs and therefore have no reference for their capabilities today. So, even if we only consider LLMs, AI is demonstrably better and faster at finding obscure security vulnerabilities in software than people are, hence the massive waves of patches for almost every widely-used piece of software. AI is getting very good at coding - for complex programs you still need a person to check its work and keep it within acceptable parameters, give it guidance with overall architecture etc, but it's getting much better at those things too. It can knock out relatively simple programs/boilerplate first time, more complex code or stuff in a less popular language might require a bit of back-and-forth - but that's true of human coders too.

I know a lot of people are wishing for the bubble to burst but it's just not looking likely in the near-term. We've barely scratched the surface of what these models can do. Progress has not stagnated; the models are still improving and they're gaining functionality every day. Just look back a few years at ChatGPT 3's output in both natural language and code and compare to what LLMs produce today - it's quite astonishing how fast and how far they've developed, and that's not even considering the massive improvements in image, video and audio generation. I don't believe there is a direct path from LLMs to General AI, but that doesn't render what we have today useless.

Obviously, the environmental impact, the impact on supply chains for memory and compute power, the harvesting of data without compensation to authors, the energy demands, the water demands, the job losses - both potential and actual - these are all real-world problems that must be addressed, but they are all surmountable.

Whether these issues are handled efficiently, ethically and equitably will be decided by human beings, not AI. Ask an LLM what it thinks the solutions to these might be, compare to the rhetoric from the tech bros and (some) politicians, and you might have a marginally less jaded opinion of the LLM.

Transparency: I use LLMs regularly for work and hobby projects, but have no financial stake in any AI companies or related fields whatsoever. AI was not used in the writing of this comment.

The only evidence of this is an email from Epstein... to Epstein, written after the pair had fallen out. That's why there's no trial; because there is no reliable evidence. You can choose to believe it or not, but don't just state it as fact.

Yeah, but "mathematical abstraction accelerator" probably wouldn't receive as much funding.

"Hey Gemini, can you go over all these documents and remove any PII from them. Now, don't be tempted to remember any of the data; just forget what you've seen, remember: we're placing our trust in you. Love ya, Gem. xxx"

Look, I'm the last person who would normally defend Apple, but let's not pretend that exposure on the Apple store, access to the Apple market and association with the Apple brand is nothing. By uploading your app to the Apple store you've got potentially millions of eyes on your product.

If your app doesn't appear on the official Apple store, good luck getting users.

Other app stores are available now for installation (well, sideloading via a web browser really) in the EU, but I imagine they will have a vanishingly small percentage of the market on Apple devices, and 99.999% of purchases will continue to be made on the Apple App Store simply because Apple users trust Apple or they have no interest or are not sufficiently informed to do otherwise.

Ok, if you get your LLM to write your papers, you're probably not going to learn much.

What LLMs actually are increasingly good at is finding the relevant information needle amongst the Internet haystack.

It's absolutely on the user to check the LLM's sources (which it should provide) and verify it's not using bad sources and/or hallucinating, but for quickly pulling together an answer from multiple diverse sources it's pretty good. I use it regularly when I can't be bothered to click through lots of Google results to find the one thing I'm actually looking for and how it relates to other things.

Do I run the risk of becoming complacent and losing my ability to search for things myself? Possibly, but unlikely. I still manually search a lot.

Do I run the risk of thinking "I know the LLM can find it if I need it again, so I don't need to remember it."? Absolutely, but on the other hand, as I get older I find myself retaining less information anyway if I'm not actively using that information daily, so having that facility is useful to me.

It's a tool. it's like everything else; if you use it to write your papers without then you'll likely not understand your topics. If you use it responsibly as a supercharged search agent, it can significantly speed up the research process - but, again, you've got to be diligent.

Good students will continue to be good students. Lazy ones will get LLMs to simply write their papers, and all we can hope for is they eventually get caught out for knowing nothing except how to prompt an LLM.

"...but, most importantly... we're making the world a better place... through slimmed-down, open-weight AI models."

What strikes me as odd is how many of these pirate operations seem to register their domain names with US companies and host on US infrastructure. If you ran a pirate TV site, surely you'd see that as a major vulnerability? TFA even says that it appears domains registered outside the US (Russia and the Netherlands) remained active.

Lol... what? Tell me you don't understand Google without telling me you don't understand Google.

All that query does is tell you all the sites that Google has indexed under that address, which is obviously a small subset of all the links that exist.

You genuinely think they had a page with all the share links (which have random GUIDs) in a big list ready to be crawled? No, and Anthropic said as much:

So, these links escaped by other means. Does Chrome automatically send pages it hasn't indexed to Google when they're viewed (which would be dodgy, but also not surprising) or did people actually post them in places where Google could see them? It's simply one or the other.

It's not possible for Google (or anyone) to simply conjure up every single valid link - there's a reason GUIDs exist; they're globally unique.

How would that work? You think they had a page with all the share links (which have random GUIDs) in a big list ready to be crawled? Doesn't seem likely at all. Much more likely: these links escaped by other means. Does Chrome automatically send pages it hasn't indexed to Google when they're viewed or did people actually post them publicly?

Requires no education, just "vibes". All you have to do is read from a corporate script, and be open to being exploited by having no say in which companies you choose to represent until you strike it big. Success is absolutely not guaranteed, competition is fierce and simply by being highly visible on socials you become a target for weirdos. Sounds like hell to me.

OFFENCE TAKEN! Offence... taken.

Exactly. This is a non-story. The page is readable by anyone if they know the URL. If you post that URL on socials, Google's crawlers are going to find it. Dumb users do dumb things, although in this case it appears Anthropic may have reached out to Google and got them to remove (or hide) everything under site:claude.ai/share - which at least mitigates the issue for now.

You're wrong.

There wasn't one monitor that kicked this off - multiple models were affected (at least 8), nor did it only affect new installs.
Colour gamut is controlled by the driver and profiles can be configured on the monitor itself. It does not require an app to function.
Calibration requires a separately-purchased calibration sensor. No LG monitors ship with an included calibration sensor. Without one, the LG Calibration app is useless.
The speakers and microphone have their own driver. They can be controlled from the monitor itself, or through standard Windows interfaces. They do not require any apps to function.
The LED lighting can be configured from the monitor itself. It does not require an app to function.

Sit down, be quiet and stop spreading blatant misinformation.

The LG Monitor App Installer is not the driver - it's an ad serving/data gathering platform disguised as a utility that downloads and installs LG other apps, specifically: OnScreen Control, LG Switch, Dual Controller and LG Calibration Studio. If you don't know what those apps do, that is fine because they are completely unnecessary - the monitor will work absolutely fine without them. LG auto-install it because it makes them money.

Source: I have an LG gaming monitor with none of the shitty LG apps installed. It works fine using the built-in controls.