jwarden

Wouldn't it be possible to just fix a single activation, just set a continuous input to what the harness knows the role actually is. Models could then be trained to trust that input and not other signals about roles.

Create of Quality New here. You are probably right here. The raw ranking formula is right, but I am lumping in the additional adjustments from the flamewar filter, etc. with "penalties." I will clarify that.

Yes I should update the text on the page. Currently it says:

> This page shows stories whose rank on the Hacker News front page is significantly lower than their raw rank, indicating that they have been penalized by Hacker News moderators.

But flags, the anti-flamewar algorithm, and other automatic adjustments that aren't moderator actions may definitely be contributing.

Neat idea. I would like to see a writeup of the thinking behind it.

Invokes several ideas: the Keynesian beauty contests and self-resolving prediction markets, and the Bayesian truth serum.

As a recommender system, I like the idea of a system where you gain some sort of credit (points, reputation, etc.) for upvoting something that eventually becomes popular before it becomes popular.

I see a huge variety of topics and languages. I think one problem is that most markets are going to have very little participation. Maybe starting with a filtered selection of already popular or trending-popular videos.

That’s a lovely thought but that’s not how China works. China is not a democracy, the geopolitical goals of the Chinese government are clear, and no large strategic company acts without approval and heavy influence from the CCP.

7stems.net: a Spanish verb conjugator and method for learning Spanish verb conjugation, where irregular verbs are just verbs with more than one stem. Also just self-published a book.

This explanation seems very implausible to me. By lowering the rent by X%, and therefore reducing annual revenue by X%, you admit the building is worth X% less. But by leaving the building X% vacant, also reducing the annual income stream by X%, you and the bank can somehow pretend the building is worth what it would be if full? I doubt owners and banks actually believe this. Is there some policy that forces this?

Edward Abby wrote about southern Utah. Red rock country. Box Elder County is northern Utah. Night and day.

I think this is right. More specifically, the problem is JavaScript function colors mean Sync/Async, whereas Zig's mean Non-IO/IO. Using function colors for async is fundamentally unnecessary, whereas for I/O it is fundamentally necessary. You should be able to define a synchronous function that calls an asynchronous function. But it makes no sense to define a non-IO function that calls a function that does IO.

EDIT: with the exception of doing IO on a freshly allocated, in-memory buffer that doesn't escape the function call.

Here’s an article I wrote a while ago about a hypothetical language feature I call “folded application”, that makes parameter-list style and folded style equivalent.

https://jonathanwarden.com/implicit-currying-and-folded-appl...

I think we should avoid suggesting that other people on this forum are evil, even if you think their ideas and arguments are harmful.

I think sometimes people are so certain about their beliefs that they perceive any argument that challenges them to be evil, bad faith trickery. But I think the best way to respond to these arguments is simply to give compelling reasons why they are wrong (and not why the person giving them is bad).

Otherwise, some people will be mislead by these bad arguments and you will have done nothing to help but say “don’t listen to him he’s evil”, which is not very convincing really.

What if the truth is that something is a lie?

Promoting truth and opposing lies are the same thing.

This writeup is very useful simonw.

But the title of this HN post is extremely misleading. What happened is that OpenAI rewrote the mission statement, reducing it from 63 words to 13. One of the 50 words they deleted happens to be "safely".

It’s like being a manager.

Capitalism just means private citizens can own the means of production (e.g. start a business, buy stock) and earn a return on investment. It doesn’t mean only the rich must get richer. It means anyone who saves and invests their money instead of spending it gets richer.

However capitalism is perfectly compatible with a progressives taxation system such that the rich get richer at a lesser rate than the poor get richer.

Right, so for a subset of functions, a language could implement an extensional equality test operator `==` for two functions by calling the functions for every possible input. It would be prohibitively slow for some functions, but correct.

But for other functions, even that won't be possible.

The point is that functions and arrays may be practically different. You can always do an `==` test on the contents of two arrays, but you can't do the same for two arbitrary functions.

Yes that’s a good point.

Reading the news and being informed are two separate things. Being an informed citizen, the kind that democracies need to survive, also requires 1) being informed of history and 2) understanding issues in depth.

People who consume a lot of news tend to have very shallow understanding of a broad range of current events. Worse they tend to be passive receivers of news instead of active seekers of information with intent to understand the world.

As a result, they are very susceptible to manipulation through selection of what makes the news they tend to consume. They become passive pawns in political power struggles.

One case where a function is often not substitutable for an array is equality testing. In a language where any two arrays with the same elements in the same order are equal ([1,2] == [1,2]), the same cannot always be true of two equivalent functions. That is because extensionally equality is undecidable for arbitrary functions.

Arrays and functions may be mathematically equivalent but on a programming language level they are practically different.

Ouch! I don’t think that’s fair.

It sounds like his professional life or personal interests naturally being him in contact with a social circle that isn’t fulfilling socially. Doesn’t mean there is anything wrong with him.

I say, look outward! Intentionally get involved with other social circles.

Listen to them.

Assuming a self-similar 80/20 Pareto wealth model (a fractal Lorenz curve), the headline number generalizes as:

    Just x%   | % wealth held | × poorest half
    ---------+---------------+----------------
    20%      | 80%           | 4.00x
    4.0%     | 64%           | 3.20x
    0.80%    | 51%           | 2.56x
    0.16%    | 41%           | 2.05x
    0.032%   | 33%           | 1.64x

This reminds me of the philosophical distinction between "sense" and "reference" introduced by Frege.

https://www2.lawrence.edu/fast/ryckmant/On%20Sense%20and%20R...

But you will never guess that the latest tik-tok craze will last another 50 years, and you'll never guess that Saturday Night Live (which premiered in 1075) will end 5-minutes from now. Your guesses are thus more likely to be accurate than if you ignored the information about how long something has lasted so far.

It does seem like common sense that they would be linked. But there is also research:

https://thesocietypages.org/socimages/2008/02/06/correlation...

I haven’t heard that statistic before. And the formulation seems imprecise? Does continuously beating the market mean that every single minute your portfolio value gains relative to the market?

I am working on Community Notes for Bluesky: https://github.com/johnwarden/open-community-notes