Not a newt

No surprise there. Abbott is, after all, a piss baby.

If them having more capacity means they’ll have a better lifetime under constant writes (because the same sector gets overwritten less often), then they’ll be good for dashcams and local storage for security cameras.

The implication of the phrase is not that trust is variable depending on the subject’s or object’s physical abilities. Its meaning is more that one doesn’t trust the other person’s motivation, so the trust only goes as far as one’s own ability to affect the other’s figurative position (“throwing” them.)

OP must be québécois.

What if this is the new normal? I reckon newer models won’t suddenly stop discovering flaws, right?

My (non-expert) impression is that what has been recently discovered is a particular category of flaws that had until now not been considered, and that’s why we’re seeing such a surge in found vulnerabilities: everyone and their grandmother is fishing across Linux subsystems for the novel pattern. The optimistic outlook is that there will be a saturation point where nearly all flaws matching this pattern have been found, and the incidence of new reports slows to a trickle.

Just wait a year or two and Google will kill Gemini on their own.

If the “literally any modern website” part is a hard requirement, then I would suggest to go for a fork of modern mainstream browsers with strong privacy/adblocking features. The justification for this is that browser engines can only get so light in terms of complexity and still support modern sites, but sites themselves can be made to be less resource demanding on the browser by selectively blocking unwanted elements. Adblocking is the obvious, blocking unwanted JavaScript would likely be the next best bang for the buck, but even clearing the cache after each session can make the browser feel faster if your bottleneck is memory/cpu instead of the network.

woo

I owned both GameSir and 8bitdo controllers. All my GameSir ones have fallen apart (stick drift, trigger broke off), while the 8bitdo ones are still alive despite my kids’ best efforts.

My career is already impacted by others using it, whether I use it or not. Those who rely heavily on LLMs produce worse and larger code, and those relying on it heavily are not the best in the first place. It’s turning -1x developers into -10x developers on account of them causing additional cognitive load on everyone else.

As for me? I don’t have FOMO. If I’m right and it’s a bubble that will collapse, then I’ll be better suited to weather it. If I’m wrong and LLMs are all they’re cracked up to be, then I will be able to get up to speed quickly.

Before Apple released the iPhone, when BlackBerry was Research In Motion, they were the biggest enterprise mobile device company. Everyone wanted a RIM job at that time.

Anthropic has every incentive in that court filling to provide as high a number as possible, because the bigger the number the better of a contract they can expect to get out of it. The fact that they didn’t means that they couldn’t (without lying).

Daniel has been quite vocal about his views on AI slop reports, but he’s also been honest about how some AI systems have been able to identify issues in the curl code, ranging from documentation drift to actual vulnerabilities. It’s not that Mythos isn’t finding vulns. It’s that Mythos is not noticeably better at finding them than other tools (LLM or non-LLM), unlike what Anthropic are claiming.

“Hey, listen!”

Congrats to all the nominees!