False dichotomy.

Centraur is when the neck of an quadruped animal gets its neck replaced by an human upper body… So the giraffe centaur would have neither a long human stomach, nor a long human neck.

It is reductionism to say that something isn’t a giraffe, if it doesn’t have a long neck part. Shame on you! 😉

In terms of analysis, I’m annoyed at Cohn here. This isn’t something we as individuals have control of. Her saying people individually have to make the difference is like saying you individually have to make the difference regarding climate change by making different choices, like recycling.

I understood her differently. I understood that she advocated into making it possible to leave platforms, saying that it currently isn’t. She said the people are the victims here and often don’t have a choice.

People cannot leave platforms because each platform is like an isle, and leaving it means losing connections to other people. It that sense they are locked-in, by social pressure.

This is is a natural monopoly which, gives social media companies so much power and prevents newcomers (like the fediverse) from joining the market.

Making the current social media companies less important, for instance via privacy laws, means people can connect and stay connected to other people via other means. It makes it easier to just leave twitter or meta, if they don’t like it there. Instead of being peer pressured into right extreme politics, because the algorithm decided that it gets more engagement when surrounding thrm with nazis.

She made it clear that replacing an dictator with another dictator that censors differently is bad, so she made a point against bluesky and for Mastodon and the fediverse.

(Sadly ehe wasn’t given the opportunity to fully complete her arguments though.)

Well… It cannot even search databases without errors.

LLMs just produce plausible replies in natural languages very quickly and this is useful in certain situations. Sometimes it helps humans getting started with a task, but as it is now, it cannot replace them. As much as the capital class want it, and sink our money into it.

That would actually be the wrong thing to want. In an ideal system trust would always begin by the owner of the hardware, where possible, not the software or vendor they decide to trust.

First the person that bought the system should take the ownership by overwriting the previous owners keys, and from there start signing the vendors key, they decide to put their trust in. Because it is important that the system is trustworthy to the end user/owner first.

Any anti-cheat mechanism relies on not trusting the person that owns the hardware, and why would that be good?

Well, it shouldn’t be carbon neutral… It should used to get carbon out of the atmosphere and into a less damaging substance.

Carbon capture does not replace getting rid of our dependency on burning fossil fuels.

We wouldn’t get back the same amount that we are burning anyway. So this approach is worse, because dumb people think it would save us, without us changing the way we produce energy.

It is worse, because it is a distraction from actually doing something.

It really depends on the purpose. Sometimes you can hide stuff in unexpected places when there isn’t much interest for other people to find it, or if they don’t even know about it’s existence.

Also sometimes it is good enough to just delay the discovery of something for a while, because its value after a certain time diminished completely.

So, I would argue that sometimes security by obscurity can be useful. But I agree that it generally shouldn’t replace proper encryption.

Yes he is pathetic and he is a bully.

And too many people think bullies that just take what they want are ‘strong’.

deleted by creator

Tons of people in STEM fields dreamt about it for years.

Not a native English speaker but “dreamt” is past tense, so they stopped dreaming, implying they stopped because now we have “AI”.

Prediction and pattern recognition is not general AI. This is just what LLMs and image generators do. They find plausible continuations starting from a noise to better fit the disired outcome. They don’t have real contextual knowledge about a domain. To them everything is just numbers that can be manipulated until they fit better. They don’t just instantly know the correct or incorrect answer because of a deeper understanding on the matter.

Well… We still don’t have AI. We got language and picture generators…

True. But most good stuff isn’t a solution for everyone. It takes real effort to escape vendor-lockin. Bigtech made sure of that.

If something is too simple to set up or requires no set up, or comes from a for-profit company, but doesn’t cost anything, then it always suspicious.

I am just saying that the issue is not with passkey itself, but the individual implementations and that google/twitter/etc. is pushed towards regular users.

Critiquing passkey because vendor-lockin is like critiquing HTML for allowing ads.

I use them with bitwarden and a self hosted vaultwarden. If my phone breaks, no issue. If my server breaks, I got local backups… Keys are stored encrypted in a postgres database for which I have access, if I need to restore it. No lock-in issue or risk of loosing access when one or two devices break.

A better, well defined API for password managers to insert login information to the site compared to text boxes.

I store the passkeys in my self hosted vaultwarden, they are a good replacement for auto inserting random passwords via text boxes.

I echo the criticism of the term ‘sideloading’, before it started to mean just installing software, I assumed it meant using a separate device or software on the side, like a PC with a debug interface or memory inspection tools, to inject custom code into a running system or software.

Similarly to preloading libraries into games or other software to replace functions in order to change or enhance the game or software. For instance used with script extenders or game mods. There it is ‘pre’ because the software is not running yet. ‘Side’ would be on running software.

But installing applications (the distribution doesn’t matter) is in no way side loading.

And I really hate that the press or whoever picked this term up from apple or google and ran with it without question.

And now, because that term is so strange and useless in that way, its definition keeps getting changed into whatever the industry needs in order to squeeze out more money and personal data, while taking away the freedom and rights of the owners.

Apart from questionable quality of the result, a big issue to me about LLMs is the way it substitutes human interaction with other humans. Which is one of the most fundamental way humans learn, innovate and express themselves.

No technological innovation replaced human interaction with a facsimile, that way before.

I think humanity is really slowly being replaced by LLMs.

Presentation and simple, but stupid and wrong ideas, are preferred over actually researching and understanding situations, isolating the underlying issues and working on ways to resolve or at least lessen them.

Just like LLMs, fewer and fewer people seems to care about a deeper understanding, and more about if the stream of words look ‘good’.

In every article on records about of food preparation, they never say how much of it is eaten and how much of it is thrown away.

I would necessitate that all or a large percentage of it needs to be eaten for the record to count.

But how can they sell priority boarding then? Just think for one minute about the poor airline companies! /s