Ha ha, the days I spent in front of my computer are finally paying off!!

Wholeheartedly agree with TFA. People who claim Firefox should go all in and block everything and return no data to advertisers need to explain how Firefox should continue to fund development.

Touch controls are becoming increasingly common in airplanes and then backed up by mouse cursors. Flight critical controls still need to be backed by physical hardware but stuff like route planning etc is now almost entirely touch based. For light sports aircraft’s even flight critical stuff can be approved as touch controls.

Aimbots dont need to do a lot to provide advantage at the highest level. Moving “perfect aim” from 1x1 pixel to 3x3 pixels, but with 33% probability would provide a huge advantage and be undetectable.

Modified assets cannot be verified unless you lock the system down, like an Xbox. On a PC? No way. You can combat it by sitting in ring 0 (which is what anti cheat software does) but you couldn’t just check some checksums.

In terms of aggregating data and spotting something like see-through walls, there isn’t the statistical method to discern between great intution built over years of playing the same map and having see through assets.

I used to work in AAA game development, across most of low level (graphics, networking, memory, assets etc) so unfortunately I know this problem is nigh on impossible to solve unless you have a locked platform.

They do do a lot of verification on the server side. Since unreal introduced their server-side-lagged-approval networking model, all local movement and most shooting can be retracted by the server.

But what a ring 0 level driver is looking for is other software, like aimbots, modified assets (transparent walls, custom shaders etc) etc. To be able to detect all that it needs to be level 0.

What I would trust more is if Microsoft acquired one of these companies and worked across the industry to root cheating out. Giving some random company ring 0 access feels completely off to me.

No. That doesn’t make it right though.

Man the particle effects of that game were unreal.

I’m not asking you to prove anything. I’m saying I haven’t seen evidence either way so for me, it’s too early to draw conclusions.

I’m not saying humans and LLMs generate language the same way.

I’m not saying humans and LLMs don’t generate language the same way.

I’m saying I don’t know and I haven’t seen clear data/evidence/papers/science to lean one way or the other.

A lot of people seem to believe humans and LLMs don’t generate language the same way. I’m challenging that belief in the absence of data/evidence/papers/science.

I mean I have an opinion too; what I’m seeking is evidence.

I mean I have an opinion too; what I’m seeking is evidence.

In this case I think it’s the DMA they’re butthurt about.

May I introduce you to our friend and saviour, the GDPR regulations?

I think I know enough about these concepts to know that there isn’t any conclusive proof, observed in output or system state, to establish consensus that human speech output is generated differently to how LLMs generate output. If you have links to any papers that claim otherwise, I’ll be happy to read them.

Well, brains are a network of neurons (we can evidentially verify this) trained on … eyes, ears, sense of touch, taste, smell and balance (rewarded by endorphins released by the old brain on certain hardcoded stimuli). LLMs are a network of neurons trained on text and images (rewarded by producing text that mimics input text and some reasoning tests).

It’s not given that this results in the same way of dealing with language, given the wider set of input data for a human, but it’s not given that it doesn’t either.

They’re only introducing it to pets because it requires very few approvals compared to human consumption. Once it’s worked for a few years (assuming) and their production systems have ramped up, seeking human approval will be much easier.

Pet food is a stepping stone.

The article makes the valid argument that LLMs simply predict next letters based on training and query.

But is that actually true of latest models from OpenAI, Claude etc?

And even if it is true, what solid proof do we have that humans aren’t doing the same? I’ve met endless people who could waffle for hours without seeming to do any reasoning.

VPN to your home network (wg-easy is the easiest way to set this up) or change the public port of your server to something the work network will allow.

Yes I get your point. Some software can run without a large income stream, on a volunteer basis.

You’re using that fact to say that Firefox also can. And if you care to look at my profile you’ll see I’ve argued time and time again that Mozilla is an overblown organisation and should be slimmed down to a couple of hundred, working solely on the browser.

I doubt, however, that you can build a modern, up-to-date browser on a volunteer basis.

How many full-time people do you think it takes?

Each to their own; may I suggest our friend and saviour Google Chrome? 🤣