Mordisquitos

I am ironically disappointed to learn that I have always been using bubble wrap correctly.

The value you are getting is not from the music as an artistic product, but from the social connection and entertainment it offers your friend group. The meme songs and lasting memories you are getting are fungible with regards to other entertaining and emotionally salient creations you and your friends may share with each other, and not with regards to other pieces of music.

I only clicked on the article with no intention of reading it (no time), but rather out of morbid curiosity as to why on earth anybody would need to be told that LLMs should absolutely not be used to generate passwords.

> [...] Despite this, LLM-generated passwords appear in the real world – used by real users, and invisibly chosen by coding agents as part of code development tasks, instead of relying on traditional secure password generation methods.

Jesus F'ing Christ. I hope to have time to read the whole thing later.

I agree that using *asterisks* just feels wrong for italics, and are much better suited for bold. However, surely the _underscore_ is just perfect to visually indicate underlining?

As for /slashes/, which would visually be perfect for italics, the only reason they conflict between italics and filepaths is the fact that in both cases they are expected to be used at the beginning and end of a word. Maybe using an unnatural placement of )parentheses( could have worked as a non-conflicting indicator of italics.

Good luck unsubscribing from the Internet. Our dependency on it is the final boss of dark patterns.

An old web surfer paddles past a pair of young social media users. “Morning youngsters! How's the internet today?”

The two users look at each other, puzzled for a few moments. Eventually one leans towards the other's ear and whispers “What the heck is the internet?”

I suspect that, in practice, what many enthusiastic advocates mean by “learning AI” is actually “learning to need AI”.

In other words, the aim is to get kids used to using AI as soon as possible, so that they do not learn the skills to function without depending on it.

I really hope so. Now I must peruse all ToS that I have agreed in the past to ensure that they have an equivalent clause. I hope I'm not contractually obliged to keep using some random website or whatever for the rest of my life.

Indeed. Just knowing that the subjects who followed through with the shocks were less likely to obey the rules could be interpreted in many ways, some invalidating the results of the experiment, some just suggesting a mechanistic explanation, and some making the results even more concerning.

* Did the subjects who went full voltage stop caring about the "learning" protocol because they realised it was all fake? Then the conclusions of Milgram's experiment are invalid.

* Did the subjects who went full voltage make more mistakes because they were more anxious and fearful of the experimenter? Then underlying fear might be a mechanism for blind obedience, and further research would be interesting.

* Did the subjects who went full voltage just enjoy electrocuting the dude so much that they stopped caring about asking the questions correctly? Then blind obedience is the least of our worries, widespread sadism is much more concerning.

Maybe we can do a meta-Milgram. A group of junior researchers are tasked with implementing what they believe to be a Milgram experiment, and while performing it the subjects (actually actors) start faking psychological distress in response to having to shock the completely fake learner subjects.

One of the researchers feels guilty from the apparent panic attack his subject appears to be going through, so he excuses himself from the experimental room and approaches the lead investigator who's watching on CCTV from outside:

“Professor, this subject is really suffering from their belief that they are electrocuting the learner. I believe this is unethical, can we stop please?”

The professor replies:

“The experiment requires that you continue.”

Interesting. If we can assume the experimenter's failure to enforce the rules was mere clumsiness or incompetence, rather than an indicator of underlying intentional manipulation of the experimental conditions à la Stanford prison experiment, this can be interpreted in many different ways.

The (eventually) disobedient subjects were better at respecting the experimental process they were given than the "obedient" ones who went all the way to the maximum voltage. Why was that?

Could it be a sign that the disobedient subjects were on average more concentrated on the task at hand (smarter? less stressed? better educated? more conscientious?) than the ultimately obedient ones, and therefore were more likely to realise they were "hurting" the alleged learner and stop?

Or could it be that the obedient subjects were more likely to realise there was something fishy going on, suspecting the "learner" wasn't really being shocked, and thus were paying less attention to the learning rules?

Or was it, as the article suggests, that the obedient ones may have shut down emotionally under pressure to follow through, and their mistakes are the result of that?

Or were the obedient ones more likely to be actual sadists, who were enjoying the shocks so much that they didn't even care if the "learner" didn't hear their question, giving them a greater chance of shocking them again?

Unfortunately I think the Milgram experiment has become so entrenched in popular culture that there's absolutely no way it can be properly repeated to explore these questions.

What you say is correct, and that is why I was referring specifically for what's necessary for "features" (the term I used for "phenotypes" to make it more HN) to become ubiquitous. It is a general rule of thumb in evolutionary biology that the more diversity is observed in a given population for a particular phenotype (e.g. hair colour, height, blood group, etc.) the less relevant it is for fitness within its observed range. When a phenotype is strongly selected for in a given population (e.g. bipedalism, opposable thumbs, the ability to speak) it soon becomes dominant and there is much less diversity.

As to your example about, for instance, neck length during abundant times, that follows the same rule: during abundant times neck length simply does not matter for fitness, therefore (all else being equal) there can be phenotypic diversity in the population.

One caveat though as to how a given phenotype may become ubiquitous without favourable selection is of course genetic drift[0], given a small enough population which is isolated for a long enough timeframe. Eventually that phenotype may become selectively "advantageous" inasmuch as it is no longer compatible with alternatives, and individuals from the isolated population who this phenotype can no longer have successful offspring with individuals of a different phenotype, resulting in speciation. That's what I meant with regards to a "make nerves on important places generate more pleasure/pain in brain" genotype being incompatible with a "have more nerves on important places" one. A hypothetical hybrid creature would be a mess.

[0] https://en.wikipedia.org/wiki/Genetic_drift

But surely we should expect some basic sanity checks on published data? This isn't some petrol stations being placed in the middle of a field due to minor typos or bad rounding, or some petrol stations' prices being listed as all 1.00 £/l out of laziness, or even a case of all unknown locations being listed as 0°0'0" N, 0°0'0" E by default. What the author reports appear to be mistakes which should be rather trivially detectable on input.

Not a dumb question. The shortest (and at a glance unsatisfactory) answer is because it works, and therefore it evolved that way.

Going in detail, first consider that for a feature to be evolutionarily selected for two things have to be true:

1. It must increase the fitness of the organism that carries it, i.e. the likelihood of its carrier having descendants as compared to non-carriers ( or be a side effect of another feature that improves fitness enough to be a net positive, etc etc )

2. It must be inheritable (and, in sexually reproduced organisms, mutually compatible during embryonic development).

One such a feature has reached dominance in a given population, as long as it continues to be important for fitness it cannot really be deprecated in favour of an alternative from scratch, even if that alternative is arguably better.

That's why, for instance, vertebrate ocular nerves connect to our retinas on the inside of our eyeball, resulting in us having a blind spot. Cephalopods, on the other hand, evolved their eyes independently the "reasonable" way, connecing their nerves from behind the eyeball. There's no way a vertebrate could mutate from scratch for its optical nerve to connect to the retina from behind without causing absolute mayhem in embryonic development. Our hacky solution for the blind spot? Let the brain hide it in software.

Going back to your question, some spots of the body being more sensitive than others became critical for evolutionary fitness long before nervous systems were complex enough to generate conscious qualia, let alone enough for them to be consistently involved in decision making. Furthermore, mapping of specific nerves to intensity of feeling on the CNS would imply complex hardcoding of something which is much easier to solve with "this place important, have more nerves", and maybe would even conflict with the fitness benefit of a CNS with enough neuroplasticity to learn anew during the development and lifetime of an organism.

So, in summary, the solution of having more nerves where it matters is simple, good enough, and has no reason to be rolled back in favour of a radically different alternative.

Indeed. It is very disappointing that they chose that as the opening paragraph of their "Why" section, without even making the attempt to relate those points as to why Palantir in Europe would be bad for European citizens.

As someone who strongly supports European digital sovereignty and eliminating dependency on the US, I'm frankly very tired of so damn much of the activist discourse around these issues revolving around US-centred topics. Yes, sure, Gaza is not the US, and the US-Israel war with Iran is bad for Europe, but those are damn well not the reasons we should say no to Palantir.

If the Israel-Gaza conflict hadn't reignited a couple of years ago and thus Gaza wasn't on everybody's minds, and if the Iran attacks hadn't (yet) happened, should we then have nothing to say as to why we don't want Palantir than it's provision of services for internal US immigration policies? Maybe I should be grateful they haven't also listed Palantir being involved in period-tracking of American women in the wake of the reversal of Roe-vs-Wade.

Jesus Christ, won't the most vocal pro-European activists please stop making everything about US talking points, and start being able to take a stance from basic principles and our own interests?

Indeed. My comment about Google Maps (and Waze, etc.) exacerbating Alzheimer's was mostly in jest and an exaggeration, but I'm absolutely certain that there is a statistically significant proportion of society whose current overuse of "AI" chatbots and short video consumption will result in cognitive deterioration if it continues at their current rate.

To be clear, I'm not implying that usage of LLMs or viewing potentially unlimited video shorts is necessarily bad, nor could I guess the threshold from which doing these things becomes cognitively damaging in the long term. All I'm saying is that such a threshold exists and some people in society are surpassing it.

I'm amazed that wasn't taken into account! Many years ago, in the final year of my Biology degree, I did a paid summer internship at an Evolutionary Biology lab here in Spain, assisting in a project where they were researching relationships between metal ion accumulation (mostly zinc) and certain SNPs (≈"gene varieties"). A lot of my work was in slicing tiny fragments of deep-frozen human livers and kidneys in a biosafety cabinet over dry ice.

The reason I bring this up is because the researchers had taken the essential precaution of providing me with a ceramic knife to do the cutting (and platic pliers), to eliminate the risk of contaminating the samples with metal from ordinary cutting implements.

That some research on microplatics did not take into account the absolutely mental amount of single-use plastic that is involved in biological research, particularly gloves of all things, boggles the mind.

And/or will we see an increase in Alzheimer's disease amongst Google-Maps dependent users? Maybe we will see a bimodal split in both directions.

There is also a very good TV comedy-drama film of 1979 about some characters working towards passing the Knowledge exam. The film was, of course, called The Knowledge:

https://en.wikipedia.org/wiki/The_Knowledge_(film)

It's available on YouTube (potential geoblocking notwithstanding):

https://www.youtube.com/watch?v=oSAAB1ZmudY

The argument is that Metaverse, as presented by its billionaire promoters as a paradigm shift, even though the public hated that vision, was a flop, therefore AI as presented by its billionaire promoters as a paradigm shift will be a flop, because the public hates that vision.

That fragment of the article is not blaming AI for that. It is criticising Marc Andreesen's literal response in this 2021 interview: https://niccolo.substack.com/p/the-dubrovnik-interviews-marc...

I don't see the author implying that AI is a fad that will go away. I see her stating that "AI", as envisioned and sold by many tech oligarchs to their audiences, is not the inevitable future.

Congratulations, this is a brilliant resource. You have done one of those countless things which I often think about doing, but my utter lack of follow-through and other distractions make it a fantasy. I cannot wait to clone the repo and explore it.

As to what can be done with the data, maybe one interesting step could be a graph-database regarding laws which reference other laws or the definitions that they depend on?

Local authorities in Spain do have the authority to enact their own law-ish regulations, which are called 'ordenanzas'. For example, if I remember correctly, motorbikes are allowed to park on the pavement by default in Barcelona unless a sign says otherwise, but it is forbidden in Madrid unless a sign explicitly allows it.

I think local government in Spain has at least as much authority as it does in the UK, maybe more, but almost certainly less than it does in the US.

I may be wrong, but I think autonomous community legislation is not published in the BOE itself (the Official State Gazette), but rather in each of their corresponding official gazettes (e.g. DOGC for Catalonia, BOCM for Madrid, BOA for Aragon, BOJA for Andalusia, etc.).

I wouldn't go as far as assuming the author is projecting, but the last paragraph of the article is indeed aligned with your second point:

> Many guardians of rock authenticity still complain that today there are plenty of people who buy a Ramones T‑shirt — maybe at some big multinational chain — who wouldn’t be able to recognize even one of the band’s songs. But the truth is that neither the Ramones themselves nor their heirs ever cared about that. In that sense, Arturo Vega’s work was just as important — if not more so — than the band’s first album.

Change the URL from instagram.com to imginn.com. There are browser extensions which will do it automatically.

Even better, what about piping the different AI responses through another LLM to provide the summary? That way you save yourself the time and effort of reading all the different AI responses.

You could even pipe the final summary directly to your email/IM client and save yourself the copy-paste.

Even better, reply only with the prompt that you would have used, not the resulting text. Don't even run the prompt through an LLM.

That results in a shorter and more concise message, and the original sender can choose to use the prompt you provided on their favourite LLM from the start.

I'm curious, why would you use an LLM to translate French to English? Why not use a dedicated translator such as DeepL, which will not only save you tokens/energy, but will also be much closer to your personal phrasing?

> no innovation allowed unless it interoperates with other products that don't exist yet

Products that don't exist yet... so, future innovation? No innovation allowed unless it incentivises and streamlines further innovation? Count me in!