chbint

Claude Code was used to do the reverse engineering.

Not hard to see how an AI agent could achieve something similar even as a step towards some innocently established goal.

Poor security + hacker-like capacities for anyone using an AI agent.

What could go wrong?

I suspect his diagnostic is pretty accurate, though. The bitter lesson came up when deep learning was already mainstream. The text discusses how that happened, and it can be the case that convenience beats accuracy. Accuracy is an epistemic value, but current AI is largely driven by market values. If accuracy manages to get along, great, but other than that, market-laden convenience reigns. Commercially, it is often more convenient to even change the world in order to make it easier for our models (consider how we're willing to create special places without pedestrians or human-driven vehicles for autonomous vehicles as a "solution" for their shortcomings).

I like RSS and I use it, but this sounds like wishful thinking. Even the amount of human produced content is just too big for one to be their own curator. We have those few authors or sites we keep up, but other than that we must rely on external help, such as HN or an agent.

Their argument is not sound, but it is informative paying attention to what they consider "evidence" for AGI. A nice instance of a problem that seems peculiar to AI: it tries to define both its target phenomenon and how well it is doing towards it.

I confess it was disappointing for me. Their main claim seems to be that thinking comprises pattern matching and pattern completion--allowing them to say that LLMs do resemble something we humans do-- but that's essentially the idea behind the connectionist movement from the 1980's - the one out of which current DNN models came from. Perhaps a friend of 1960's symbolic AI would be unhappy with that claim, but there are not many of these around anymore (Gary Marcus is misrepresented as one such, but his view is that models should be hybrid, not purely symbolic).

Nowadays, the question about whether LLMs are "actually" doing something similar to human thinking revolves around other dimensions, such as whether they rely on emergent world-models or not. Whether such world models would require symbolic reasoning or not is a different matter.

This has interesting implications for representations within AI models.

You're basically arguing that the constitution is on equal footing with or inferior to the other laws. That's just backwards and to insist on this borders on gaslighting.

No. I'm making the Law-101 claim that, according to the Brazilian constitution (as well as most democratic constitutions worldwide) you can't rely on freedom of speech to bypass the law. Pretty much everything you said up to this point relies on this fundamental misunderstanding.

The constitution is the constitution. If other laws contradict the constitution in any way whatsoever, they are wrong and invalid. I don't even want to hear the argument, they are unconstitutional and they should be literally deleted from the law books. And they would have been, if these judges were doing their jobs correctly.

There's a hierarchy within the constitution itself. Fundamental principles of the constitution come "before" constitutional norms. Indeed, constitutional norms must be interpreted under the lights of the constitutional principles, and any interpretation that doesn't follow them is unconstitutional. Isonomy (equality), for instance, is a principle. The right to speak freely is a norm. That means, among other things, that your right to speak freely can't be used to harm isonomy. Abuse of economic power is a way to to harm isonomy within elections (for you can use money to boost your reach, effectively "silencing" your adversaries by flooding your discourse everywhere, making the voter's decision to be a function of your money rather than about your political ideas and proposals).

Therefore, your whole reasoning about what should be the case is pretty much backwards, for you're implying that norms should come before principles.

Whenever one wants to think of oneself as capable of teaching judges (or any other professional), how to do their jobs, it is always a good idea to study how things actually work.

By this point it's more than clear that we're never going to agree with each other.

Indeed, for I agree with the Brazilian constitution and you don't.

Qubes is a good approach to an OS, but it's Xen security, not OS security, and I'd rather run a secure OS other than Fedora or Debian on Qubes.

Not sure what you mean by "Xen security" in contrast with "OS security". Qubes is an OS. Though a lot depends on your threat model, if you have high security needs, Qubes is likely to be your best companion.

Anyway, another reasonable choice is Kicksecure. It's the debian-based OS underlying Whonix (Kicksecure is focused on security and Whonix adds its privacy/anonymity setup on top of it). You can use Kicksecure as a VM within Qubes, by the way.

https://www.kicksecure.com

I honestly don't think you'll be able to get very far unless the team is already onboard with the plan to switch towards a more asynchronous culture. If they lack the motivation, they won't bother improving their writing.

The tricky part is that they might be interested in the results you promise, but still lack motivation. It's common for someone who's interested in losing weight not to be very motivated to do it themselves. They don't want to lose weight themselves, they want to "be slimmed down by someone else". You may face a similar difficulty. That's part of the reason why changing a culture is so hard.

I know that's not a direct answer to your question, but I needed this context to say this: I think whatever tricks and tips you can come up with yourself are more likely to succeed. That's because you're already familiar with the specific needs and the specific difficulties people might face when handling the most frequent and repetitive issues.

Rather than thinking how to get fuzzy improvements on people's overall writing skills, perhaps you could try to focus on suggesting specific solutions for specific problems ("hey everyone, I've noticed that when handling X people usually forget to tell p, q and z. So let us agree on using this text structure '1) p; 2) q; and 3) z; whenever handling X"). I think that, by accumulating lots of small tricks like these over time, you'll be able to go further. Going bottom-up seems easier than trying to change things top-down.

I'm not sure this makes sense to you, for I'm not familiar with your concrete situation, but I hope it helps somehow.

Your understanding of censorship is fundamentally distinct from that of the Brazilian constitution (BC) because you clearly detach freedom from legal responsibility. In BC, freedom of speech is bounded by the law. That's why you can't (e.g.) say you're going to harm someone without facing the legal consequences.That's also why you can't use words to incite a (virtual or real) crowd in which you're influential to take illegal action against someone (and one can do that without being explicit). Thus, in BC, censorship is what happens if you prevent someone from speaking something they are legally entitled to. In stark contrast, your examples were basically of how speech could be used to bypass law limitations (e.g. abuse of economic power is a well-defined example of illegal activity).

The difficulty in seeing "where is it in the constitution" is likely related to viewing Law texts as a kind of source-code in which everything must be carefully defined and every possibility must be explicitly stated. No constitution work that way. No one could. It may contain some very specific rules, but in general it is an articulation of principles that must be observed by legislators while formulating laws. One of such principles is that our freedoms cannot be detached from our responsibility (i.e. the possibility of being held responsible). Law works that way because one can't predict in advance every kind of situation.

For instance, if I ask you whether you agree that right to life is a fundamental value, you would probably say yes. I assume the same goes for freedom of religion. But what would you say of a father that refuses to do blood transfusion in his newborn son for religious reasons, thus leading to its death? As responsible for the newborn, is he entitled to act this way? Which value should prevail? Freedom of the father to act according to one's religion or the right of the newborn to life? Distinct communities can give distinct answers (e.g. ancient Japanese would perhaps pick honor over life). The point is not whether there's a clear answer or not. The point is that you can't expect to find an "if-then" statement explicit about it in the Constitution or even in laws in general, and that you can't expect that whatever works in this specific situation is fully generalizable to every other situation in which these two values conflict. Applying the law requires interpretation of both the text and the situation in question, and there's a whole discipline dedicated to how that's done (hermeneutics).

The same goes for values such as freedom of speech. Given its intrinsic connection to responsibility, its application is always bounded by the law, and the application of the law is always bounded by the specific situation at stake. That's why we can and should tolerate people believing weird things that go against our best scientific knowledge (juridically, no one cares if you believe that the earth is flat), because we can easily keep them under check using arguments, rational discourse, etc. However, if it becomes part of something illegal like attributing a crime to someone else without evidence (e.g you're using your influence over flat-earthers and their typical suspicion of NASA, to make them believe that the NASA is supporting your political adversary in an attempt to fraud the elections), you must and will be held accountable, (that why it's not really about "judges deciding what's fake"; rather, it is about how one is (ab)using a given discourse.

Like I've said before, Brazil lives under the rule of a democratic constitution built after much fight against real dictatorship and real censorship, not the rule suggested by you here (which, again, absolutely no country in the world lives by). You're free to disagree with the basic liberal and democratic principles grounding the Brazilian constitution, but whenever you and the constitution disagree, bear in mind that it is the constitution's point of view that's going to prevail.

Judges were aware of what the producers themselves claimed the content to be and how the content was being promoted. There were clear references to the elections and to people involved in the elections. Moreover, It's not like members of the electoral court decided to bother the producers out the blue. There was a denounce that the producers were abusing their economic power during the election (which is illegal). Evidence and hearings involving their lawyers were conducted, and that's where and how the judges became aware of everything.

What you have, thus, is a scenario in which 1) the producers were already being investigated for electoral misconduct; 2) they were known supporters of B; and 3) they were boosting and promoting B.'s campaign material in social media disguised as "news" and "documentaries" (which is a way of trying to dodge the accusation of economic power abuse). The case in question is just an acute one.

And even in the face of all this, authorities didn't outright ban the release, but delayed it until after the election (about a week), effectively preventing misuse without imposing censorship.

Because there's an ongoing investigation/legal process and you need to prevent something bad from happening. Typical examples are: the risk of you running away to another country, the risk of you jeopardizing the investigation in any way (say, by threatening or harming key witnesses), the risk of you getting killed by accomplices who are on the run, and so on.

I was about to make a very similar comment. I won't say I'll never switch to neovim, for a lot depends on future vim/neovim development, and unexpected things happen.

But I do agree that vim's stability is priceless. It's been years without any need for major changes in my vimrc, and without any trouble with the plugins I use.

I'm sympathetic with the author, though. Whenever you need to change, finding an alternative that "just works" always makes things easier and you can quickly get back to being productive. I'm not so sure that I wouldn't go down a similar path if the vim ecosystem collapsed.

Where in the constitution does it say that you can engage in censorship of any kind, let alone political?

Brazilian law does not consider preempting someone from shouting "fire" in a crowded theater a case of censorship. AFAIK, no democratic country would, for all recognize that freedom implies responsibility. Same goes for those trying to shout, without any evidence, "B's adversaries ordered his killing!" (or something similar) 6 days before the election day, even though they are trying to make it look like a "documentary".

I'm using the same logic that allowed US citizens to publish and export cryptography software by printing source code in a book. This is technology was literally export controlled for national security reasons. (...) And they used free speech to publish the source code (...). So don't compare distorted brazilian notions of free speech to american ones. They sure as hell have a lot more free speech than we do.

Beautiful story. You should tell Snowden. Maybe he'll realize that he didn't need to go into exile in Russia.

I couldn't care less what the goal of the documentary was. (...) And I don't care for the judge-king's censorship of "misinformation" either. I'll judge for myself, thank you very much.

But the Brazilian constitution does care for both these things.

That's why it requires special care during the elections in order to prevent abuse. What's at stake is a principle you find in every liberal thinker since modern times, and that grounds most (if not all) democratic constitutions worldwide: If there's no fire in a crowded theater, one can't shout "Congressman X started a fire! Run!", incite people to leave in a hurry and later claim that "I was just manifesting my political opinion, people were free to ignore and judge the situation for themselves", as if you were not expecting their panic and the risks associated with crowds in panic. You're responsible for whatever ensues, and if this kind of behavior can be preempted, it must be. Or so thinks pretty much every democratic country in the world, not just Brazil.

If we judge by the rule you mention, this would be censorship. That would mean there's probably no country in the world that could be considered democratic (even the so called "absolute" US freedom of speech is something of a myth, for there are lots of decisions from the US Supreme Court that would be deemed "dictatorial" according to the criterion you present here).

Bottom line is: Brazil lives under the rule of a democratic constitution built after much fight against real dictatorship and real censorship, not the rule suggested by you here (which, again, absolutely no country in the world lives by). You're free to disagree with the basic liberal and democratic principles grounding the Brazilian constitution, but whenever you and the constitution disagree, bear in mind that it is the constitution's point of view that's going to prevail.

Similar experience here. It seems like we can learn the a pattern of distinctions pretty quickly. That's telling.

I believe this requires context:

1) The producers of the documentary in question are a politically active group and well-known supporters of Bolsonaro.

1.1) They are also known for producing (and earning money from) content that spread misinformation, conspiracy theories and the like.

1.2) They always could, and they still can, produce and disseminate this kind of morally questionable content without being disturbed. They were never subject to a priori censorship, for we are not living under a dictatorship anymore.

2) In 2018, during the presidential campaign, Bolsonaro was the victim of an attack (stabbed). The suspect was arrested red-handed.

2.2) After that, a thorough investigation was carried out by the federal police (our FBI, so to speak) that concluded the suspect had mental issues and acted alone. There was no one "behind" the attempt.

2.3) Bolsonaro's personal lawyers, who followed the investigations, saw no elements to question nor to require further investigation, and in the criminal sphere, that's the end of the story.

2.4) However, as we could all expect, Bolsonaro politically exploited his attack. Up to this day, he or his supporters occasionally claim or imply that "this politician" or "that organization" is behind the attack he suffered. None of these claims can be grounded in the investigations carried out by the Police, and he never presents any evidence, not even circumstantial ones, for any of these claims. Long story short: he won the 2018 elections.

2.5) However, in 2022 there was no ongoing investigation anymore. Therefore, if he wants to point the finger towards someone, he has to do it by ignoring the conclusion of the investigations.

3) In 2022, while trying to get reelected, 6 *days* (not months) before the elections, the producers of the documentary in question (whose name is "Who ordered Bolsonaro to be killed?"), tried to release it online. The obvious goal was to exploit the attack politically in order to help Bolsonaro's reelection.

3.1) During the elections, as in many European countries, Brazil has specific rules designed to prevent economic abuse and fight disinformation that could cause a harmful imbalance in the electoral contest. You can't, for instance, accuse your adversary (or people from his campaign) of committing a crime without evidence (specially if the crime in question is something like ordering a murder) a few days before the election in the hope that people in shock vote for you. However, political supporters and campaigners use lots of well-known techniques to try to bypass these rules. One of them is to present Campaign information in the form of a "documentary".

4) That's why the electoral authorities preempted the producers from releasing the material 6 days before the election day as they planned. It could be released freely the day after the elections. And it was. It is there for anyone to see since then. That's what happened. Nothing like the heavy prior-censorship to which every journalist, artist and citizen was subjected to during our military dictatorship a few decades ago.

I hope this helps those interested in understanding Brazil's recent turmoil.

Seems logically equivalent. If you have the right to express your opinions (no qualifier here, so they can be popular or not), that means no one can do anything to you (unless you commit a crime, of course). Perhaps one could argue that "speech" encompasses more than "opinion", but then the issue would become terminological.

Anyway, Brazil has freedom of speech in the very sense you've mentioned here. Unpopular speech is not a crime.

Sure, you're actually charged with a crime. And the kind of limitations you talk about do apply.

Even in situations where you could be held in jail, there's a tendency to let you go unless it is impossible to prevent you from jeopardizing the investigations by any other means. For instance, if the only real worry is that you flee to another country, you might have your passport confiscated rather than being held in jail. Likewise, if the worry is that you can use your influence to make others do stuff for you (stuff that jeopardizes the ongoing investigations, I mean), then you might remain at home, under surveillance, and so on.

Most Brazilians agree that censorship is wrong. The problem is that "censorship" is a vague word.

We lived in an actual military dictatorship until 1985. A dictatorship that engaged in real hard prior-censorship. Music, news, and pieces of art were all subject to a military collegiate body that would decide what could and could not be published.

What's going on now is very, very different. Brazil, like most European countries, thinks that if you commit a crime through what you say, you can and must be held accountable. No one is being prevented from expressing their opinion.

What we're facing here is a distinction between US and BR law (actually, US is the exception world wide, for Brazil law is closer to what you would find in Europe on this matter).

In Brazil, it's not a crime to say what you think. But it is a crime to falsely claim that someone has committed a crime. This is especially serious if you are influential on social media and your statement, even if false, is likely to generate dangerous reactions from your followers.

Indeed, but no one is doing that.

In Brazil there's what we call "preventive custody". If you're caught committing a crime, and if there is a risk that you could jeopardize the investigations (by eliminating evidence, threatening or influencing witnesses, etc.), then you are held in custody until the investigation is concluded.

I don't believe you would find something very different going on in any other democratic country.

Indeed, but useful nonetheless. Solving it may be a challenge for a while, and deep fakes generated before a solution becomes broadly available will remain detectable with this technique.

No need to postulate platonic forms. All we need is the idea that there are real patterns to be mapped. The idea that distinct nets can share a representational space is around at least since Laakso And Cottrell published their "Content and cluster analysis: assessing representational similarity in neural systems" in 2000. If you look for "representational similarity analysis" you'll find more research about it.

Indeed. The concentration of such a huge power consumption might make it easier (policy-wise) to take compensatory measures.

There's indeed a nice trick to be learned from cognitive science focused in biological cognition: the mind is embodied and embedded. Which means, roughly, that it is not portable. It doesn't store things like "glass at position x,y" but only "glass is at a small movement of the hand towards the right". Consequently, whatever gets encoded only makes sense within a given body and only inasmuch as it relied on its environment (with humans, that includes social environments). The good news is that, despite being not portable, this reliance on physical properties might be a step in the right direction, after all.