IanCal

It is. It’s also often not necessary at all. You can’t do things with people’s data without either getting consent or basically having a good reason to. I like the ICO pages (uk regulator) for explaining a lot of things like this.

If I’m shipping an item to someone I don’t have to ask them if I can keep their address for long enough to send them the item. I do need their permission to use that data to send them marketing though, or sell it on. If you have to legally keep records for X years that’s fine.

Keep only what you need, for the time you need to keep it, in an appropriately secure way.

It’s the type of rag the article is about however, it’s still very relevant.

They can include the protocol, it’s that regardless of what’s put in there it’ll use https, which is explained with examples both in the following sentences and then in more detail in a linked doc.

> For repositories on GitHub, GitLab, and Bitbucket, a specifier now names a repository rather than choosing a transport. github:owner/repo, owner/repo, git+https://…, and git+ssh://git@… all resolve through the host's canonical HTTPS URL, and the lockfile never records an SSH URL for those hosts. To reach a private hosted repository over SSH, configure the machine with git's own URL rewriting:

    git config --global url."git@github.com:".insteadOf https://github.com/

> pnpm shells out to git, so the rewrite applies to all of its git operations. Unknown hosts keep their exact URL, SSH included, and a URL with embedded credentials is kept verbatim and never resolves to a host archive. Details in How git dependencies are resolved.

Drop me an email if you’re after a quote.

Or I could automate it. It’s so much of every post these days.

I expect the complaints would be fewer if it was a smaller thing, or if everyone had their own version rather than it feeling like one company deciding if you should be able to use a large fraction of the internet.

I’m going to write a filter to remove comments that are just bitching about whether something is AI written or not. Seeing “slop” now starts to give me m$ vibes

What’s the problem with that? Seems to make sense to me with the example and it’s important to point out. I’d expect an llm to explain in more detail to be honest.

> RAG is basically good old information retrieval with LLMs doing the querying.

No - rag is doing search before you call the llm to give it context from some corpus like your helpdesk articles.

It depends where you are, some areas have very good bus services and other terrible ones. Almost everywhere has been forbidden from having a more centralised approach and so most places are a mash of random companies with very different approaches and reliabilities.

It’s not hard to get started, it’s a case of adding small amounts of randomness.

If you have, say, a long poll then kick off all users due to a deploy or error (or a broadcast message) then you can have a situation where you’ve got a huge clustering of connections at 1 minute, which spreads very slowly out as real life issues give you jitter for free. You can avoid this or at least return to normal much quicker by adding some jitter.

It might happen if all your users back off at the same rate too, if the clustering causes a bunch of errors. Error -> lots reconnect 1 minute after -> fail -> lots reconnect 2, 4…

More likely to occur in cases where there’s a way you can have people all connecting at the same time - synchronisation to a real world event is one case and then connecting again at the same time after.

How do caches work across models? I would have thought that was very model specific - if not I’ve really misunderstood what’s getting cached.

Could you imagine a scenario where from warning to complete evacuation takes more than two days? Evacuating a whole area is a hard task, particularly once you start looking at more complex problems (elderly, prisons, hospitals).

— Spoilers, if you’ve not read this and yet are still reading this comment chain

I guess the grooming may start earlier, it’s not really discussed and it’s a bit ambiguous as to when that started

> but relents almost immediately,

Physically, at the time, though this is something they’ve argued about for 6 years.

It’s not a passage I particularly care for, and it could have been less explicit but then I’m mixed on how that would change the story. I’ve just finished it and it’s certainly an interesting sci-fi story.

Does it provide a strange contrast to the other things?

There’s disgusting horrors inflicted by terrible people but to the willing (but are they only willing due to what was done to them before, by PI?).

Deliberate and horrible suffering of a single person caused for minor gain.

There’s calm destruction of trillions, unfeeling, for protection out of a measure of harms and trying to protect one set. There’s the good goal or perhaps just self indulgent goal that led to that too.

There’s the deliberate killing of trillions with glee at breaking things for one persons view of humanity. There’s the deliberate doing of this by the original catalyst, but with less clear direction. Perhaps a lack of logic and more boredom?

And yet this, done for improving survival chances of a group, feels over the line. I don’t disagree that it is quite disgusting but I do find it interesting at that being my reaction given what else has been done up to this point.

s/model/engineer

Oh I mean LLMs are the opposite, they’re not the same blind optimisation thing weirdly watching you with no core understanding. They’re great for this kind of thing.

Tbf if I don’t enjoy that aspect and it saves me a day or so per year that I can use for consulting instead it makes sense to do that trade. Two and a half perhaps including taxes.

They didn’t say fibre didnt matter. They disagreed they should care more about fibre than any other.

Recommendation algorithms are, frankly, weird as a user. Something watches your tiny interactions and, with zero concept of anything happening, gets surprisingly good at predicting your next interactions.

It’d be like eating and having an alien with no concept of food or taste staring at how much your facial muscles change and making notes. Then recommending places to you based on the other people they’ve been studying.

> how is AI (or a human for that matter) supposed to know what “woodworking-appropriate music is”?

Well someone is going to be concentrating, but it’s a long period, with loud activity. maybe you’d guess at decently active music, not relying on careful complex listening, probably fairly steady rather than sudden dramatic changes.

Maybe it could save previous playlists and what you’ve said about them, or ask questions like “so like, really calm or techno or what?”. Adding just a bit of common sense makes an enormous difference with this kind of thing.

I think you’re right, my base assumption is that the models can code and can fix bugs, and can code more in parallel and faster than humans at a lower cost.

If Google are tackling lower value bugs with AI the number is in a way inflated compared to some utility measure (fixing a smaller number of worse bugs could be preferable) but it’s still things fixed.

This is the research taste part - often they can be good but human experts are really good at this (also you know your codebase, without the prompt these models have no other background).

Then being able to suggest several things to try and have them go off and build, measure and tweak is hugely useful in my experience.

Also things like making custom visualisations for comparing changes.

I can say a portrait doesn’t look like a person even if I can’t do better, googles product range is an absolute mess and using it is a nightmare - I don’t know how to fix the company but I know the outcome is bad.

What’s it a prediction of?

The question of agency and the question of income are different. The former - we do things manually that can be automated for fun all the time and make entirely arbitrary weird things to do for no value. However some find it hard to move from work to leisure.

For income - who knows?

> Google acting like a normal but competent company that's just chugging away,

Chugging away not releasing products or releasing ones worse than their competition typically. Not sure that’s what a normal competent company should be doing.

IMO that doesn’t sound so much like prediction any more.

It’d be prediction if it’s “predict what would come next in this text sampled from distribution X”.

But what’s it predicting if we’re looking for new useful outputs? It’s finding a distribution that’s useful, and generating tokens, but it’s not predicting what comes next in a known sequence.

It’s always been nebulous but that was fine because we were so incredibly far away from it. We never really planned to be close to it figuring out the edge. Some have it at human or beyond for all tasks, but then rarely touch on “one human” or “all humans”.

Personally having been in AI since before deep nets, systems have been incredibly narrow for decades.

Classifiers on images were battling with ten classes in 2010. Imagenet had 1k classes and people were getting half of the things wrong then and that was frankly amazing at the time.

And they only did images, only to known classes, only with very specific inputs.

Text classifiers only did a few classes usually and mostly threw all the words together.

The most advanced things I saw in the late 2000s were struggling so much to make general systems that the most general ones were still incredibly limited and bad at those things (we had a robot learning to play games that you showed it). Things like asking a thing for a book and having it parse the sentence, identify what was needed, that it didn’t know where it was but that was knowledge another human had and asking them - that was impressive yet also limited to very small sets of interactions.

The idea of a machine getting sarcasm, even if mostly built for it, was wild.

General meant capable of a broad range of tasks without retraining.

To me we have agi. It’s general, and it’s good enough to be useful.

We have those laws. You need consent regardless of how you get the data. That’s they say “we value your data” (a phrase you can read in two ways) and ask if 1368 partners can have it for various uses.

Browser fingerprinting is not cookies. Cookies are a very specific technical thing.

GDPR is not about cookies, and what data do you mean? The data about how you use the site, tied to anything they can identify you with?