hangsi

To aid curious readers, the cell biology term for this (many genes to do X) is pleiotropy. The Wikipedia article [0] conveniently includes a discussion on its relevance in autism.

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

It sounds like you might be a physicist who disagrees?

This is from the interview on Hot Ones (released August 5, 2021): https://www.youtube.com/watch?v=yaXma6K9mzo&t=816s

> Sean Evans:

> I think a scenario lots of viewers can relate to is sitting on the couch on a Friday night, going through the streaming services, cycling through the movies and thinking to themselves "they're not making movies for me anymore". As somebody who's been intimately involved in movie making for 30 years, what are the macro Hollywood conditions behind that sentiment?

> Matt Damon:

> Well, so what happened was the DVD was a huge part of our business - of our revenue stream - and technology has just made that obsolete. And so, the movies that that we used to make: you could afford to not make all of your money when it played in the theater because you knew you had the DVD coming behind the release and 6 months later you'd get a whole other chunk - it would be like reopening the movie almost.

> And when that went away, that changed the type of movies that we could make. I did this movie "Behind the Candelabra" and I talked to a studio executive who explained: it was a $25 million movie. I would have to put that much into print and advertising to market it - what we call P&A - so now I'm in $50 million. I have to split everything I get with the exhibitor, the people who own the movie theaters, so I would have to make $100 million before I got into profit. The idea of making $100 million on a story about this love affair between these two people... Yeah, love everyone in the movie, but that's suddenly a massive gamble in a way that it wasn't in the 1990s when they were making all those kind of movies - the kind of movies that I loved and the kind of movies that were my bread and butter.

There is a mechanism for this, internationally usually named some variant of a Law Commission [0]. The idea is to look for laws that are technically in effect but can rarely or never be applied. For example, the UK Law Commission boasts a repeal of 3000+ acts in its time [1], such as repealing rules for conducting slave trades that were made obsolete in the 1800s but not repealed at the time.

In addition to the sibling comment's mention of the Congressional Review Act for agency oversight, there is a US Office of the Law Revision Counsel [2]. It has an official website [3] which is beautifully old-fashioned, but looks to be purely a resource for accessing the letter of the law and doesn't recount its volume of repeals in the same way.

None of this matters if the insane or counterproductive regulations are deliberate and desirable for the current lawmakers, of course.

[0] https://en.wikipedia.org/wiki/Law_commission [1] https://lawcom.gov.uk/repeals/ [2] https://en.wikipedia.org/wiki/Office_of_the_Law_Revision_Cou... [3] https://uscode.house.gov/

Divided we stand,

United we fall.

The almost-wrong simplification is that a nonlinear medium changes the wavelength of the light that passes through it.

If you can control the nonlinearity, you can control the wavelength change and so change properties such as the angle of refraction to change where the light goes (like in a rainbow/a prism, where the red light refracts more).

> can we not expect writers to just keep it brief?

In essentially every situation, you cannot expect readers to read.

I recall from my own visit that the electrical transformers are supposedly original. So, the National Museum of Computing justifies calling its Colossus a rebuild rather than a replica, since it is made with some original parts.

This is excellent. I wonder how deep the roots of pre-20th century computing systems go. Babbage, Lovelace and the Difference engine are well catalogued, and I have seen a Jaccard loom in a museum with my own eyes.

What comes before this that isn't a history of mathematics, aside from the abacus? If this search is broad enough to include the topic of this article and Luca Pacioli's briefly mentioned double-entry ledgers from Italy, then I can imagine systems from all over the world where commerce flowed or administration ruled: similar systems must have existed in China and India, and I have heard of the Quipu system in the Andes that functioned as a digital storage medium for thousands of years.

How many modern components of information systems are reinventions of past ideas, rather than upgrades?

This seems untrue, especially for reddit?

Using the 1/9/90 split [0] for creators/commenters/readers, it seems farfetched to suggest that reddit accounts (which benefits readers making an account to curate subreddit subscriptions) can't follow this pattern where many legitimate human users do not comment often.

[0] "The 1% Rule", https://en.wikipedia.org/wiki/1%25_rule

You are right. The grandparent post ironically uses word this in a cheapened, shallow way when they can use it freely in their own writing.

You could argue this is the very sort of activity they were criticising when they posted! We are all vulnerable.

Tell that to the Pierce-Arrow company: makers of the first official cars for the white house, but they didn't survive cash flow problems from the great depression. Meanwhile, Ford survives.

https://en.wikipedia.org/wiki/Pierce-Arrow_Motor_Car_Company

You would think, but there are many "sea lochs" along the coast that are salt water inlets (lagoons?), and Loch Long is one of them. I think it's a grey area (and perhaps the West coast of Scotland in general is too).

Yeah, there should be some sort of world wide epoch shifting environmental change about 65 million years ago to show for it.

This seems to have an obvious counterexample?

The name of the Streisand effect is from exactly the situation of a photo nobody cared about of Streisand's house from decades ago. The fact that it can be superficially referenced is evidence of its longevity.

Nazis, traditionally. Typing it out makes the comparison feel weak.

https://en.wikipedia.org/wiki/First_they_came_...

Interesting to consider GPUs as the coal of the AI revolution.

This is worth it for the mental image of heaping them into a boiler fire by the shovel load alone.

The common method for choosing the next output token for an LLM is sampling from a Boltzmann distribution. If you have seen the term "temperature" in the context of language models, that is a direct link to the statistical gas mechanics.

Yes, but then it needs to be stored, and the price effects might encourage even more production. Previously this has led to events like the devastating Wisconsin butter flood [0] ("A firefighter reported flames 300 feet (91 m) high", "It took about twenty hours to contain the blaze, and eight days until the fire was officially out").

Not that rice would do anything so nefarious, but more to point out that large scale purchase and storage is not a trivial task. I can imagine the scenario where some bureaucrats ran the numbers and concluded that it is cheaper to just give the money to farmers instead of running the purchase/storage scheme.

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

Arguably Crowdstrike's loss was due to a flaw in their product rather than PR.

Disney's product (though not their core, I would argue) allegedly killed this poor woman - however it is the lawyers' behaviour, not the death itself, that is an additional PR liability for Disney; death from an allergy is tragic, but could have potentially happened at any restaurant in the country. Only Disney (and a select few other large corps) could pull this particular bad act in defence.

As an aside, the entire line of argument from Disney is an absurd legal fiction. No reasonable person reads terms and conditions, and so they should not be bound by the terms. I hold a weak hope that this case is bad PR for the practice as a whole that raises the profile of this injustice.

Clues must be related to the semantic meaning of the cards, so irrelevant clues are technically banned. So, a "sausage" strategy would be illegal on many boards.

Yes, I agree completely - I realised something was off in an espirit de l'escalier way not long after I posted.

This is great news for my proposed box redesign!

This is an interesting view of how random number based security is compromised for economic practicalities (though the meanings of "security" and "compromised" might be overstretched here).

For completeness, I wondered how many cards would be required to give the complete set of patterns.

If we number the positions in the 5x5 grid such that the top row has positions 1-5, the second row has 6-10 and so on, the grid positions can be converted to a sequence and we can use the permutation formula to find the number of arrangements. To account for rotations, we can divide the final value by 4 since every arrangement can be rotated and is therefore valid.

Of the 25 cards, there are 7 white, 8 red, 8 blue, 1 black and 1 double agent that can be red or blue, also deciding which team goes first. We can treat this final card as one of a kind, then double the formula output to account for cases where it is swapped to the other team.

Permutations of a multiset has a standard formula [0] that calculates a result from these values (rolling in the double agent factor of 2 and rotation division factor of 4):

25! / (7!8!8!1!1! * 2) = 946,551,177,000

(edit: as pointed out, this is 9 times too large as the double agent can indistinguishably replace each of the other 8 cards - a corrected value is 105,172,353,000)

This is (edit: still) more layout cards than have ever been printed across all production runs of Codenames, and would probably not fit into the current box size.

[0] https://en.wikipedia.org/wiki/Multinomial_theorem#Number_of_...

The argument I can imagine is already around (like for many potential AI applications): even if you know the law, such as if you are a lawyer, you always get representation because judges and jurors are prejudiced to rate self-representing participants worse.

I can easily imagine the same (unprincipled) dynamic applying to an AI lawyer.

3 throws per second is the rate of action required to perform juggling literally. The implication is that communicating between the miss that causes the fall and the event of dropping the ball must happen in less than 1 second. Translating back to the metaphor, communicating that the ball is being dropped before it hits the ground is a task as hard as the juggling itself.

"Beautiful. Unethical. Dangerous."

So says Morgan Freeman's character Lucius Fox in 2008 in The Dark Knight[0].

The rest of the tech imagined in that scene is plausible today too, considering the density of WiFi/5G and research demonstrating the potential for its use as passive radar [1]. That paper metions a cooperative base station, but I am wondering if there is any value gained in knowing exactly what the traffic is (such as some of the intelligence community does) in modelling how the waves propagate and performing an even more passive observation.

[0] https://www.youtube.com/watch?v=IRELLH86Edo

[1] Samczyński et al. 2021 https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=966...

I think the parent comment was referring to something else.

In the paper the tasks are only completed by GPT-4V. For a valid scientific investigation, there should be a control set completed by e.g. qualified doctors. When the panel of experts does their evaluation, they should rate both sets of responses so that the difference in score can be compared in the paper.

Neural networks have two different compute costs: training and inference.

These are roughly analogous to compile time vs runtime for compiled programming languages.

Training is in general a more intensive task. However, in an ideal scenario training is run once and inference is run millions of times, so the lifetime cost of inference is bigger - this is why it might make sense to optimize for intense.

It's from The Elements, by Tom Lehrer (the final lyric of the song).

https://youtu.be/qyNkrlwKs7c

This isn't really copyright - more the microeconomics of losing market share/control.

Specifically, I think they are less concerned with (say) specific Excel code leaking than with the knock on effects of a cheap perfect substitute.