"Canonical's announcement doesn't use the term 'AI' anywhere in it."

STTs are not LLMs, but they are AI that use CNNs and RNNs.

The models Canonical mentions by name are Whisper, Parakeet, Nemotron, and Qwen3-ASR (https://github.com/canonical/myna/blob/main/docs/architecture/Myna%20-%20System%20Architecture.png).

"If you have a good keyboard, you can be far more efficient with it than with voice input."

-1, provably false
https://www.medrxiv.org/conten...

"Except, of course, for the front part, which was weirdly aerodynamic"

The command module is common to several other craft.

" (nowhere to keep fuel, for example)"

In the corners between the rear engines (4 engines in a diamond pattern) there are spherical tanks.

" good visibility of the ground "

Except you can't see the ground anyway since the nose extends several times further than the depth of the cockpit and the base of the forward windows is at or above eye level.

"giving the connector that level of admin permissions"

An exploit would not require *being given* that level or permissions but rather *obtaining* that level of permissions, i.e. chaining exploits. If Google can prove that it is not possible that there will ever be an exploit that would give that level of permission which could then be chained then Google would be justified at calling it just a misconfiguration. (Google cannot prove that.)

Even then, I would say it's a design flaw if you allow a configuration that one one would actually want.

"removes functionality"

AMD never claimed that the functionality was there and neve promised that it was going to be there.

I suppose you could work some estoppel argument.

The question that any kind of "recognition" facility can answer is given a specific identity and some metric what is the likelihood that the metric comes from some rando and not the specified identity. Given enough identities, any confidence level will get you at least one match.

This is why you never use the same sample to calculate a p value for multiple factors, and also why biometric passport gates *do* work (you are only comparing one metric against one identity).

At a 93% confidence level, you'd expect at least one match for every 15 identities tested against the photo.

What I suspect happened is that they ran every photo for which they had identities available and returned the highest confidence. That's a guaranteed loser strategy.

There are three kinds of tabloid narrative conflict:
Florida Man against Nature
Florida Man against Florida Man
Florida Man against self

-1, illiterate

This story is reporting on the journalist's actual experience paying someone to perform this modification and actually watching them in person performing the modification. It is not mere copypasta of a troll post.

In the end, it was found that the original marshmallow test didn't deconvolve impulse control from trust in authority. If the students don't trust that AI won't be used against them, e.g. faulty AI checkers used to accuse them of AI use, their non-AI skills not valued on the job, etc., then it is rational for the students to use AI.

In 2004, $500 million in GPU orders were transferred from Tesla to X. This year, SpaceX acquired X. Now they're being hired out.

The only AI that wasn't racist enough is also the only AI that has more GPUs than it can use.

"Nothing is more capitalist than trying to maximize your revenue."

a) The motivation under capitalism is maximizing profits, not revenue.
b) John Deere isn't trying to maximize their profits either; they're trying to minimize the profits of others. That's comes from a belief that all transactions are zero sum, which is definitely NOT part of capitalism.

Consider the scenario where you capture all of your customers' surplus, or all of your suppliers' surplus, plus one cent. Your customers not suppliers have no reason to say in business. At that point YOU are no longer capable of staying in business. Enlightened self-interest drives you to keep your customers and suppliers profitable.

However, the median businessman has no idea how capitalism works; at the 90th percentile they are actively opposed. At that point they are not capitalists, they revert to mercantilism.

If you stop measuring the ocean, it will go away.

A token is actually well defined. The problem is that it isn't exposed.