aleph_minus_one

> Cute. Of course, beyond basic deal breakers, neither do people know their preferences

Scott Alexander disagrees:

> https://www.astralcodexten.com/p/in-defense-of-describable-d...

(see also the follow-up article https://www.astralcodexten.com/p/highlights-from-the-comment...).

Important quote from this article concerning your point:

"Here are some things that are so obvious they sound like cheating:

Unless you’re bisexual, you probably care if someone is a man or a woman

Many people care whether their partner prefers polyamorous or monogamous relationships.

Many people want a partner of the same religion and level of religiosity.

Many people want a partner who matches their preference about whether to have children, and how many to have.

…because in practice, people end up very closely matched on some criteria

Only about 4% of marriages are between Democrats and Republicans

Only about 3% of high school dropouts marry a spouse with a college degree, compared to more than 80% of PhDs.

90% of whites marry other whites; 80% of blacks marry other blacks.

Husbands and wives’ social classes correlate at about 0.8

If 96% of Democrats are marrying non-Republicans, it seems like Democrats must have a strong preference against marrying Republicans, and ought to value having information about someone’s politics before they date them. Realistically, this underestimates the level of political sorting; I don’t think I’d be a good match for an extremely woke person, even if we were both technically “Democrats”."

---

> Specifically, interests, hobbies, daily routines, objectives, etc. have nothing to do with compatibility.

Of course your claim "'We like the same things!' is the compatibility signal of a naive 22-yr old…" is correct, but I am very certain that there exist quite a lot of interests, hobbies and objectives that require a partner that is "compatible" with these (and in most cases, people who have such interests, hobbies and objectives know this very well). This obviously does not mean that the partner has to have the same hobbies and interests. :-)

For example,

- there do exist good reasons why in university time, sports climbing was basically a dating market

- a person who has hobbies that are associated with semi-legal activities (which clearly does not mean that he/she does anything illegal!) like hacking in the classical sense will likely not be a good fit for a spouse who works in law enforcement

- a person who is a climate activist will likely be a bad fit for a spouse who works in the oil/gas industry

- a person who is very into internet privacy topics (think cypherpunk) will likely not be compatible with a social media influencer

> If we theoretically hit the point where it could do all of its research itself, better than a human, things _would_ be different.

If we theoretically found a way to shield or reverse gravity, things in aviation or space travel would be different. Or if we theoretically found a way to make cold fusion work, things would be very different. :-)

It is in my opinion not a good idea to invest in companies for which the feasibility of the business models depends on the capability of making science-fiction stories work.

> The problem is twofold. One, even a monopoly AI provider wouldn't have pricing power against its suppliers. Its suppliers are energy, semiconductors, and real estate. Semiconductors maybe they could get some leverage on but energy and real estate have plenty of other buyers.

Concerning the leverage on energy and real estate: don't forget that the AI companies have quite a lot of choice where to build their data centers. So AI companies have lots of opportunities to play several parties off against each other (in particular also for real estate and energy).

> From a business perspective a frontier model does not make much sense anymore if you are not a startup. Neither for Amazon, nor for Google. Their clouds need models that are fast and perform well in their agent frameworks nothing w[h]ere a frontier model excels at.

A good consideration; just one point from my side: as far as I am aware (but I may be wrong), Gemini is not known to perform well in an agentic framework.

This is no contradiction to your other claims, quite the opposite: perhaps (or even likely) Google wants to avoid that their models become a commodity in some (agentic?) application where the middleman who actually writes this application gets a disproportionate of the money that the customer of the application pays for it.

> China will always be generations behind until they crack domestic EUV

Are you sure?

--

China Just Built What TSMC Said Was Impossible

https://www.youtube.com/watch?v=Pk-w279ESHg

--

China Just Built What ASML Feared Most

https://www.youtube.com/watch?v=YiPgSm62fiM

> Then why have they been lagging behind OpenAI and Anthropic for most of the last few years, and only briefly been at the frontier?

One possible explanation: because Google is a little bit more frugal and focuses on how to make providing AI models financially feasible - combined with some willingness to burn money so that they don't strongly fall behind on their AI models.

On the other hand, OpenAI and Anthropic at least formerly concentrated on building and providing the best models that they could with concerns about financial feasibility taking a backseat.

Just to be clear: I do have the impression that by now (likely because of pressure from investors) OpenAI and Anthropic take these financial concerns more seriously, but nevertheless Google's vs OpenAI's/Anthropic's "DNAs" concerning on what to focus on differ.

> I guess if one of them hits singularity, it could in theory just wipe out all the rest

The story that some AI company might reach singularity and then "everything will be different" is another science-fiction story that executives of AI companies love to tell to justify the staggering amount of necessary investments and cash burn. :-)

> The famous theory of Dario Amodei was that AI was this winner-takes-all field where the first team to get a head start would never cede ground back.

This is the kind of story that ones tells to investors to justify the huge amount of cash burn. :-)

Why don't you simply ask the respective model which model is best for a specific task? :-)

As far as I know, this depended a lot on the individual game:

Games that had quite a bit of code that profited from the higher-level abstractions that C/Pascal provided, such as

- resource management (including dynamic asset loading/unloading)

- enemy AI

- pathfinding

- ...

of course typically contained a lot more C/Pascal code than games that were more on the "multimedia demo" side with a simpler game logic.

> it was called Cactus Data Shield.

Indeed:

> https://en.wikipedia.org/wiki/Cactus_Data_Shield

Additionally, I remember from this time (perhaps this was a very German thing) that bands and stars whose CDs used Cactus Data Shield or a similar copy protection system suddely got hated and abandoned by parts of their fanbase.

> Nope, they were mostly written in Assembly.

It depends: for the classic 8 bit home computers, games were mostly written in assembly. Later DOS games were commonly either written in Pascal or C, but quite a bit off Assembler code was often used for the more performance-critical code sections.

> Presumably this is a so-called “chicken bit”.

For those who are out of the loop:

> https://lwn.net/Articles/744840/

See also

> https://en.wiktionary.org/wiki/chicken_bit

(click on "quotations") for a list of scientific texts in which this term occurs.

> It's not every day you see someone so incredibly dismissive of the product they're selling themselves.

I believe that considering that AI is much cheaper than their service offering, and thus customers of Dymocks Tutoring and Talent 100 often ask them how much more they are actually offering for the much more expensive price has put this company under a lot of price pressure from customers.

So, I bet that Dymocks Tutoring and Talent 100 wants to change their main service offering to target parents from rich families instead who aren't so stingy.

Perhaps tell the AI that it should rather "praise" you as much for the progress that you make as a Chinese parent.

:-)

(to make it clear: this was a joke, even though I'd actually be interested whether this works)

> But I also have a soft spot for Myst.

Riven is even better. :-)

Because there exist so many schools of programming and so many deep philosophical considerations what makes software good vs not good.

Basically, software created by programmers who are more on the artisan side often deeply encodes what they value in life and what they don't.

So, finding a "good" programmer for a company means finding someone who is skilled in programming, and who has very compatible deep opinions about life and everything.

Also, many exceptional programmers are not good, obedient workers in a "software factory"; this is another consideration that you have to keep in mind.

> I'm shocked in general at the lack of open source Rapid Application Development software.

They exist:

- In particular in the 90s and 00s, Tcl/Tk was really magic for creating user interfaces for small applications under GNU/Linux.

- Lazarus attempts to be an open-source analogue to Delphi

- LibreOffice Base attempts to fill the same niche as Microsoft Access

- Historically, in the time of Gtk+ 1.x and 2.x Glade [1] was an editor for clicking together user interfaces. Well, development ceased, and under GNU/Linux there is no attitude of retaining backwards compatibility with applications/libraries that are not maintained anymore (but which might be business-critical).

- ...

But of course Rapid Application Development is much more popular in the business world which is (at least in many European countries) a rather Microsoft-centric environment. Additionally, the typical developer that uses Rapid Application Development is often a subject-matter expert in some department who also knows a little bit about programming (i.e. is not in deep love with programming), and not a programming expert who also knows about some other specialized topics that are important for the business.

On the other hand, the open source environment typically rather attracts people who really like programming.

---

[1] https://en.wikipedia.org/wiki/Glade_Interface_Designer

> Everyone should be using zstd

Why?

> I have seen senior Oracle sales reps deliberately misreading their license doc to inflate what is owed.

"I guess I can't pay so much because such a huge fraction of the contract volume goes to my lawyer."

> In a place operating with SCRUM you're often expected to be shipping things every sprint

... or you don't care about these expectations, and give a big "f*ck you" to them.

> will just define a new baseline expectation

This assumes that you care about (at least a little bit) what the other side expects from you.

> The senior you get the more you realise that the only reward for hard work is more work. Then, you learn to say no more often and set boundaries and negotiate like if you want me to work on project A, then I won’t have capacity to work on project B.

This is something that you already realise very, very early in your career. The only hard thing is how to set boundaries and negotiate without infuriating the other side and burning bridges forever.

> Someone should make the millennium problems for AI alignment so the frontier labs will actually try to solve that problem.

Rather: Someone should make the millennium problems for human alignment so the top researchers will actually try to solve that problem. :-)

> AI writes unmaintainable code - you can see that many projects don't accept it.

There also exist other good reasons why projects don't want AI-generated code, in particular

- because of unclarity of copyright status and consequences of AI-generated code

- because the project leader simply made the observation than many programmers who hand in AI-generated code care more about "getting things done" and "pushing through their changes" (possibly to boost their CV) instead of deeply caring about code quality

> No way. Mathematicians don't assume things to be true even if there are no counterexamples.

It depends: there do exist some conjectures for which there don't exist any proofs, but a huge amount of results assuming this conjecture (which is very near to assuming that the conjecture is true). Examples are:

- Riemann hypothesis [1]

- Generalized Riemann hypothesis (GRH) [2], and potentially also Extended Riemann hypothesis (ERH) [3]

- Standard conjectures on algebraic cycles [4]

- perhaps Schanuel’s conjecture [5]

--

[1] https://en.wikipedia.org/w/index.php?title=Riemann_hypothesi...

[2] https://en.wikipedia.org/w/index.php?title=Generalized_Riema...

[3] https://en.wikipedia.org/w/index.php?title=Generalized_Riema...

[4] https://en.wikipedia.org/w/index.php?title=Standard_conjectu...

[5] https://en.wikipedia.org/w/index.php?title=Schanuel%27s_conj...

> The doc won't point it out explicitly and I apologize if my explanation offends anybody, but the intention is basically leftist, state-friendly bitcoin. The authors took a look at the anarchy of real cryptocurrencies and realized they don't actually like anarchy, so instead they want to make something "wholesome" that governments would be willing to go along with (since they can tax it).

While this is correct, I want to add a point:

Many hackers are not per se opposed to taxation (how else should public services be financed), but deeply distrustful about the red tape involved (the taxation laws in Germany are insanely complicated). Basically, their attitude is: if the government wants to tax "my" cryptocurrency, it is the government's responsibility to make the taxation laws as elegant, privacy-conscious and hassle-free as humanly possible.

In other words: if the government does not create taxation laws that are as hassle-free and privacy-conscious as possible (in a very strong sense: if in terms of effort (including the time to understand and apply the laws), it is much easier to ignore the laws than to turn in your taxes, the government failed its job), the government failed its job, so it does not deserve a concession from the cryptocurrencies' developers side.

And many Germans are deeply distrustful of such cameras.

> They brag about their great "data protection" laws, and the same same day you read in the news how yet another of their universities got hacked and how some incompetent government agency leaked the sensitive private information of hundreds of thousands of their citizens.

I would claim those people from Germany who are very privacy-minded indeed have a huge distrust of these data collections exactly because they know that sooner or later someone will crack these systems.

> Hacker conference: against authoritarian government, with talk about security, privacy, opsec etc.

> Also: can only buy ticket online with credit card and real name, no onsite sell available, your name is probably give to conference administrator for checking ticket, he will probably give it to third parties or keep it for "good reasons", etc.

I am indeed quite certain that many hackers in Germany independently already devised the same thoughts ...