Well, of course they are, but so is just about everything

Great thumbnail!

Dunno (or care!) about the rest of it. Kinda strange content for the technology community…

Holy crap - get to the point!

This is a great idea at first glance - I’ve certainly had a few usb sticks come apart on removal

That’s some bob cringely like dissection - love it!

Good point - it guess it could have easily fallen out while being edited, too

The start(-up?)[sic] generates up to $2 billion annually from ChatGPT and an additional $ 1 billion from LLM access fees, translating to an approximate total revenue of between $3.5 billion and $4.5 billion annually.

I hope their reporting is better then their math…

Yeah, there’s some limits to what they could do while maintaining pace for the 0 day stuff…

Some input validations would be the most basic things they should have done years ago. I’m aware of the hashing mature vendors do of any content they download for updates or deployments. Signature checking as well, and that’s before the code is even inspected - why don’t they include their automated tests they obviously aren’t using in the update as a sanity check client-side? (I’m not aware of anyone doing this or even if it’s possible without the rest of the IDE, stack, I’m no dev)

…sorta. The complexity here is their driver is signed, but it’s also loading code from their channel file (that was all zeroed out), and it seems the necessary error checking wasn’t implemented.

I haven’t yet got to the root cause they published, this is just what I gathered from the video of a retired MS kernel dev who posts stuff.

Obviously with their design it allowed them to be flexible at the cost of playing with fire - I’m impressed they got away with it for so long, really

Wow, that’s some amazing stuff! Really looking forward to more of this globally…

You’re right - this is fucking stupid… Unfortunately, it’ll capture the business of small time shops that don’t know better

I really fail to see the problem. The open source community has long had a tendency for wacky names.

That would be awesome!

Yeah, but back then you Americans had a government. I can’t see it happening with your current circus.

The EU, however, is already looking at MS over teams monopoly practises (fucking finally!), I’m hoping edge and copilot/bing are on their radar too!

Holy shit, really?

My change controls written in comic sans are going to a new level!

It’s great to see some progressive web developments after all these years of regressive trash

I tried to follow this but my brain is fried (and it’s only lunch time!)

One thing it got me thinking about (and I was surprised by the conclusion I came to), was it’s often brought up how the training models are black boxes that are proprietary - but we all know the data was whatever public records they could scrape from the internet, be it reddit or whatever.

Such a thing didn’t exist for them to use in a licenced manner, they were innovating - so I’m naively wondering why is it a problem when they took the risk of using the data and presumably paid tremendously low wages to people to prune and train it from 3rd world countries

They still had to build the thing and pay to run it, train it and mature it. The risk was all theirs, why is it a problem that they’re now hoping to profit from that?

• Maybe they should sell their training data…

We’re upset at the greedy little pig boy spez for licencing it to them, but we did chuck all our thoughts up on the bathroom wall for all to see. It’s not like there was anything private about it.

I do like the approach of changing the incentives, but that will need regulation to force the capitalists to behave, so I guess we’ll just have to wait for the EU to form a plan.

Just a low-level filter that inspects and vetos things (think anti virus)

Awesome tool, that one. Not often we use it (and usually inside a virtual application environment), but it’s great to rely on…

It’s so ridiculous I’m not even going to bother with the article