Posts by Michael Wojcik

Re: Red-line

Eh, you know the old saying: Don't ascribe to shilling what can adequately be explained by fanboyism.

Re: Terminology

Yes, and it's a term of art in IT security.

Shall we complain about, say, the use of "facial recognition" rather than "guessing"? We could be here all day.

Re: What has immigration status got to do with criminal punishment?

At least you guys would never make an idiot president because he was funny on a TV show

Damned calumny. Trump was never funny.

Re: Monkey See, Monkey Do...

Distinguishing mechanical interlocutors from human ones is how the Turing Test is defined.

Turing's "Computing Machinery and Intelligence" is not a long or difficult piece.

Re: Turning tests

The Imitation Game (which is more than simply "talk to the computer") wasn't proposed as a test of sentience. It was proposed as a test of mechanical thinking, which is a different cognitive category.

And its force is not as a practical decision procedure anyway, but as a shot across the bows from the good ship Pragmatism. The question Turing's essay examines is "on what grounds, if any, can we argue that the external attributes of cognition are not sufficient to conclude cognition?".

That said, I agree that people serious about AGI don't think the Imitation Game is a good practical decision procedure. Robert French cataloged a number of objections to that project in a CACM piece years ago, for example.

Re: Definition

It's likely there's no "universally accepted definition" of any concept. With billions of human beings with vastly different sets of life experiences, probably all of whom hold at least some perverse opinions, universal agreement just isn't possible.

It's also not how language use and the interplay of ideas among humans work. In the best case we converge on functionally constructive intersections of understandings.

So your "universally accepted definition" bar is useless, I'm afraid.

Re: generally agree...

machines are machines

Sure. Now show that human beings aren't machines, and that human cognition isn't mechanical.

I believe we're still a long way from AGI, based on my own experience with ML, my readings in AGI research, and my experience in other fields (particularly the study of writing and the philosophy of mind). And I think "sentience" is very much the wrong metric, and largely uninteresting, and I wish the media would stop bandying the term about as if it's some sort of gold standard.

But I've yet to see a persuasive dualist argument that human cognition is somehow special and can't be mechanized. (And, yes, I've read Penrose's.) I don't even think the QECTT needs to be invoked; I think cognition is most likely conventionally computable.

Re: There is a name for this....

Yes, and at first glance it looks like it was an attack against a wallet implementation, not against the cryptocurrency itself. Not that it matters much, particularly to the people who lost (notional) money. (How much of that could be converted into hard currency or goods & services is always a question.)

Re: Interesting

There are plenty of good applications of Merkle trees already, like zfs and git. We really don't need good applications of blockchain, which is just a toy Merkle tree.

As Bruce Schneier has pointed out, we already have plenty of algorithms for append-only ledgers. Blockchain isn't novel and isn't particularly interesting.

I routinely support software that's more than ten years old, and not infrequently over twenty years old.

Not all software is Windows.

the big thing about "smart contracts" is they are supposed to be code that has the full force of a legal contract

"Supposed" by whom? I don't believe I've ever seen that claim even by smart-contract enthusiasts. It's hard for something to have the full force of law when it's not recognized by the law.

Smart contracts aren't contracts under any sensible definition of the term.

But that said, I agree that it's not immediately obvious what law or contractual arrangement might have been broken here, so it's not clear to me what criminal or civil action would be available, even if the perpetrators could be identified. Maybe something under the rather sweeping and arbitrary computer-misuse statutes that legislatures have pumped out, since violations of those are pretty subjective. Maybe a terms-of-license violation, if there's also a license agreement that attaches somehow. Eventually there's likely to be some tax-code violation, unless the perps report the income properly in their jurisdictions.

Re: I wonder how much blame can be attributed to poor code examples

Code samples? Who has time to read code samples?

I assume the developers of these apps just copy and paste from StackOverflow.

Re: Greedy bastards blatantly milking a broken system

USPTO rejects about half of all applications. That's hardly "rubber-stamping".

Re: ...and on Mac, don't dare to use emojis

How about "don't use emoji", full stop?

("emoji" is a Japanese word, and as such both the singular and plural are "emoji". "emojis" is a barbarism.)

Re: Screen Reader support

Screen readers don't have a problem with plain text, in my experience. Perhaps someone who uses them more extensively can explain why complex tables would be better. (My guess is that they wouldn't.)

Writing well is in part a design process, for any useful definition of "design". That follows clearly from both composition research and from rhetorical theory. You could also support it based on, say, linguistic pragmatics.

Re: Seems the Magic 8 Ball was right then...

Outlook (aka Lookout!) has always been a mess. Mine averages two synchronization errors a day; it especially has difficulty with the concept of deleting a message.

Around 2000 I pointed out on VULN-DEV that even if you disabled pretty much everything you could disable, Outlook would still insist on rendering Windows metafile graphics embedded in messages. Because there could never be an exploitable vulnerability in the complex metafile rendering engine, yeah? So stupid.

Of course, the fundamental problem is MIME, which was a terrible idea and we're all still paying the price for it.

Re: "One of these was Boris Johnson, but he paid the price,"

This is Alito. It's far from the worst thing he's ever said.

We're talking about the SCOTUS justice who thinks it's fine if the police murder anyone they think might be guilty of a crime, using whatever means might come to hand. He's a lousy jurist, a lousy thinker, and a lousy human being.

Thomas is arguably worse – arguing in favor of allowing corporal punishment in schools, for example – but he's not such a loudmouth.

Weirdly, so far the Trump appointees have managed to be not as bad as those two, Dobbs v. Jackson aside. Gorsuch was on the reasonable side of cases involving compelled speech and LGBTQ+ employment rights, for example; he's also been decent on tribal rights. Kavanaugh was on the right side in Garza and Trump v. Vance. Barrett doesn't have much of a record yet, but she hasn't been as stalwart a friend of business and authoritarianism as Alito and Thomas (see e.g. Hollyfrontier). We'll see what happens in coming years.

Re: I used to be nice to christians, not any more

there are even those who don't have any all-powerful boss you have to blindly obey

Sure. Sikhism, though it's technically monotheistic, rejects the idea of a monopoly on absolute truth, for example. Hinduism and Shinto give you a whole catalog of gods to pick from.

On the other hand, while Sikhism historically hasn't been used to motivate religious oppression (rather the opposite in fact), Hinduism and Shinto have. Neither monotheism nor the concept of a single supreme authority are necessary for abusing religious beliefs as an endorsement of violence.

Personally, I don't invest in untestable hypotheses (which is what, by definition, any belief in the supernatural is). But I think the vitriol directed against religion is a bit misplaced, because what history shows is that ideologies can be grounded in many sorts of beliefs, and can be harnessed to oppress regardless of the ontological status of those beliefs. Take for example the eugenics and scientific racism movements, neither of which were yoked to religion.

I do think there is something to be said for programs in rationalist thinking, even though we know humans will never be very good at it; and I even agree with Richard Rorty (against, for example, Stanley Fish) that non-substantive commitments to certain philosophical ideas can help people be better. But on the whole it's really easy for people to justify being shitty to one another, and getting rid of religion wouldn't change that.

The thing about files is that the provide pieces of storage with a purpose.

Exactly. The story of information technology, since long before there were mechanical computing machines of any sort, much less digital ones, is a story of partitioning and organizing information. Written language was invented specifically to label bullae and represent their contents. "Files" are contemporaneous with the earliest efforts to store and manipulate data.

Works like Yates' Control Through Communication illustrate the evolution of this process. If POSIX filesystems bear a certain resemblance to earlier technologies such as pigeonholes, flat filing, and vertical filing, that's because those systems did the job.

Personally, I'm far from convinced of the benefits of non-volatile RAM, variations of which we've had for ages. Keeping everything in memory is a whopping great increase in the attack surface, for one thing. Having multi-level storage and control over my working set is just fine with me, thanks anyway.

OS/400 famously had its "single-level store", where all objects were mapped into a large virtual address space. Obviously the implementation had to make use of virtual memory and paging because the hardware couldn't physically support all the data the system had access to, but it implemented just the sort of "everything's an address" metaphor that Liam is asking for. It was OK. It was not revolutionary.

There are other ways of organizing user information which make more sense for particular use cases, like Sugar's journal mechanism for children learning to use computing technology. The filesystem metaphor isn't necessarily optimal for every use case. But I'm struggling to think of a use case where single-level storage really conveys any significant benefit.

Re: I don’t have a dog in this fight

TL;DR: You're wrong.

Identifying bot accounts¹ is a very different problem from estimating how many bot accounts exist.

Twitter have explained their methodology for the latter, and they've given the Muskrats access to the data. No one from Team Musk has yet shown any evidence that Twitter's estimate is far off; we just have a lot of wild handwaving claims from Elon about it.

Beyond that, the Court of Chancery is really, really unlikely to find anything about the bot numbers constitutes a material adverse effect. "Material adverse effect" is a term of art in M&A law; it's not just "anything that gives Elon sadface". The bar for an MAE is really high, full stop.

Also, the contract doesn't say anything about accurate bot numbers. What it says is that Twitter is obliged to respond to reasonable requests for information that has a relevant business purpose. Twitter gave Musk's Minions what they asked for. They provided their estimates; they provided their methodology; they provided the "firehose" of information about Every. Goddamned. Tweet. They can't invent more information. That's it. That's the whole thing.

Musk's lawyers are throwing up a cloud of smoke, but so far they haven't shown anything that has any real weight.

¹More specifically, bots posing as mDAUs, monetizable Daily Active Users. No one questions that there are a lot of overt bots on Twitter; Twitter encourages that use case. What's at issue (well, not actually at issue, but what Musk and his lawyers are pretended is at issue and what some rather gullible observers think is somehow important) is how many accounts are 1) routinely active, 2) presenting themselves as real meat people, and 3) actually bots.

And violate a court order? Hello, contempt charges!

Delaware corporations are under the jurisdiction of Delaware, including the courts of Delaware. Musk can't just take his ball and go home.

Even if he could get away with re-registering Tesla elsewhere, the Equal Powers clause would let Delaware pursue remedies.

But in any case, selling the shares of Tesla that Musk owns does not depend in any way on where Tesla is incorporated. It depends on whether the court has power over Musk's finances, and to the extent he is personally liable, it does. (Yes, the actual agreement is between paper holding corporations but I can't see the Chancery flinching at piercing the corporate veil in a case like this.)

The Court does seem to be keen to get this done and dusted

SOP for the Chancery, which moves quickly.

If their names and signatures are not on the merger contract with Twitter, then they're not liable.

Unfortunately for them, their signatures are on other contracts connected to the merger, and the substantial analyses I've seen from lawyers claim they quite likely are liable.

Re: Core issue of ???

Twitter's bot numbers are unlikely to be considered material by the court. Devin "LegalEagle" Stone did a fairly thorough analysis of the case on his YouTube channel, and he noted that the Court of Chancery is rarely willing to admit something is a material issue significant enough to void a contract.

Stone's analysis is worth viewing. Basically, nothing we've seen in public so far from the Muskovites looks very promising. Twitter is in much the stronger position.

Venting windows will not be an option. You might think you need or want to be able to open the windows, but that's just because you're trapped in assumptions based on current designs.

The entertainment system will only stream iTunes. That's for your protection.

There will be no physical controls, just an enormous touchscreen.

Re: CRYSTALS-Kyber

They're not IBM's names. The CRYSTALS (Cryptographic Suite for Algebraic Lattices) project is an inter-organizational effort. Two of the team members have IBM affiliation, but it's not owned or run by IBM.