senekerim

OTOH if you've reached a point where you are ready to accept both 'A' and 'not A', you need to go back and re-examine your premises/assumptions (sort of a real life example of reductio ad absurdum). Related example: accepting 'the left is delusional and corrupt' and 'the right is delusional and corrupt' leads one to re-examine the validity and usefulness of this whole left-right spectrum of political thought.

I have been to lots of interviews, on both sides of the table. I find most interviewers unprepared to evaluate the person for the role, and instead exercise their own biases, stroke their egos, etc. It's largely a voodoo practice that we'll look back and laugh at as a civilization at some point..

It's definitely not perfect, but it's much better than any alternatives I have seen. Would love to take a look if there are any good proposals based on S-Expressions or other computer-era notations.

Math notation is actually pretty great and is a result of multiple iterations. Earlier forms were much worse. S-Expressions by themselves are obviously not enough, you have to introduce at least first-order logic notation, set notation, etc. It's not obvious that the end result will be anywhere near as readable as the current notation.

This looks awful to my (mathematically trained) eyes. What is the rationale behind introducing this language feature?

This is one of those cases where you can't really extrapolate from history. At least linearly extrapolate.

This is written by a person who doesn't know what he's talking about. Any time people diss Bayesian stats and name Yudkowsky in their argument (a controversial guy without a high school degree), they haven't got a clue what they are saying.

Bayesian stats is a solid mathematical theory. It is used everywhere in academia and industry, is the foundation of a lot of AI & Machine Learning research, powers many technologies that, among other things, allow you to locate this thread on the Internet and post that ignorant link here.

Please stop promoting BS.

> I haven't dismissed the conclusions, just said I disagree with many of them.

Earlier: > If they uncover anything, it is the very interesting political foundations of a group of people

That sounds like a classic dismissal to me.

--

Anyway, not to get bogged down in a silly semantic argument. Who cares if it's real philosophy? I've tried reading Heidegger when I was young and naive. If that's real philosophy, I'll pass, thank you.

If you have substantial, rigorous arguments against the existential risks posed by AI-based tech, I'd love to hear those, otherwise what you're saying sounds a lot more like an emotional and political reaction (presumably from the left end of the political spectrum).

> However, it does not follow that the uncredentialed person telling you that education doesn't matter is therefore right in the rest of what he/she is telling you.

Where did I say that it follows?

Also, AFAIK Robin Hanson is a full econ professor.

> something Sheldon Cooper would write

One might view this as a style/personality jab rather than an issue with rigor.

All I'm saying is Sheldon Cooper types have moved things forward for the rest of society more often than people realize.

Also, it would help to hear why you're so dismissive of the underlying arguments. One of their core ideas is that technology (specifically AI) can grow to surpass human level intelligence and pose a threat to the civilization. This is not a vacuous statement or a lunatic's fantasy, we have seen many other types of technology develop to a point where they pose an existential risks (e.g. nuclear weapons). If anything, the dismissal of AI-related risks reveals very clear blind spot / lack of rigor in most people's reasoning. "Because it hasn't happened yet, it must not be a problem we should worry about."

> They (and his earlier co-author Robin Hanson) also like to reduce things to signaling, basically saying that education is just to show off and raise one's status by acquiring pieces of paper, but if you really want to get things done, it's more efficient to self-study and concentrate on the really important things, skipping all this mess that is education.

You say it like it's not true. Traditional education is a mess. A lot of it is smoke and mirrors and signaling. Which is why online education is huge and growing.

I'm not here to defend the particular social movement that grew around these guys, even though they are right about some things, like catastrophic risks tied to new tech (which certain academics are quick to dismiss).

And I get the point of it not being too rigorous in academic sense. Sure, no arguments there.

The point I would like to make is that dismissing thinkers due to their writing style, personality issues, oddities, etc, can lead to very problematic outcomes. Many top thinkers in history have had childish qualities and were not well-rounded fully functional "normal" human beings. "Normal"s are too busy leading productive lives to write thousands of pages on vaguely identified future risks. The types of blind spots that these odd thinkers can uncover in the rest of the society can be very useful in improving everyone's lives. The guy who pointed out the toxicity of lead had to go through same type of ridicule by the society that was busy applying a layer of lead paint on every flat surface, including their faces. Same with many others pointing out practical risks that the society is currently blind to. Off the top of my head, other famous weirdos with odd beliefs, yet huge contributions to science/tech: Faraday, Tesla.

It's not really a cheat, as it produces the MLE (i.e. most likely) estimate under certain assumptions (e.g. errors are normally distributed, which occurs naturally if the errors are large sums of many unrelated measurements errors).

One reason is that people's innate probability engine is broken, so they can't correctly figure out what's best for them (in expectation). This applies both to individuals and to civilization as a whole.