meken

I think it has to do more with infix vs. prefix notation - infix is more natural and easier to read.

> For example, the evidence favors frequent low-stakes assessments with feedback (e.g., homework, quizzes) over few high-stakes ones (e.g., exams) – but AI is undermining practice in low-stakes settings and pushing us more toward exams.

Why would AI undermine quizzes? I took an intro to Japanese class where we would have a quiz at the beginning of each class which I quite enjoyed. I can’t see how AI would change the utility of that.

> There's no solution other than training a lot, lot, lot, lot more doctors.

Seems like having doctors emigrate from other countries would work as well.

I switched from vi readline mode to emacs recently primarily for Ctrl-r and Alt-. (insert the argument from the last command). I love how interactive they feel.

You can accomplish the same things in vi editing mode (with ? and !!) but they’re not as interactive.

I recently watched the anime Neon Genesis Evangelion, where they make reference to a Dirac Sea [1].

It seemed very magical/trippy/cool.

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

No one is saying you can’t solve puzzles that have already been solved. If you love solving puzzles, then whether it has been solved doesn’t seem all that important.

> That said, if slowing this down were possible, I think it would have happened by now. Maybe he’s hoping the OAI/HF incident becomes something the industry can rally around, but it won’t.

I disagree with this and I think the reason is well captured here:

> The idea of pausing or slowing AI has been floated as far back as 2023, and I think it made little sense back then... The AI models of those days were not powerful enough to act as agents in the world in any coherent way, and were not capable of significant deception, manipulation, cheating, or cyberattacks... Today, however, the picture is totally different.

> The fundamental problem is simpler: no one will slow down because no one trusts anyone else to slow down.

Did you read the essay?

> [Embedded evaluators] is something Anthropic is unilaterally committing to (and calls on governments to require other frontier companies to match).

Hard disagree, to the point where I'm not sure there's not some confusion at play here.

I find using indices and summations generally confusing and hard to read/follow/understand. To me, it's so much simpler to draw out the computational graph, derive the gradient with a single element (so I'm in agreement with you guys here?), see how that applies the same to every element the same, then scale up to the vectorized/matrix version.

> and you need to memorize a bunch of new rules to apply it

There's nothing to memorize though..?

Is there something called "matrix calculus" that's different than just drawing the computational graph and deriving the gradient in the way I described..?

Like, is this "matrix calculus"?

- https://cs231n.github.io/optimization-2/

If so, what is there to memorize here?

I bet it can be decomposed quite nicely though. At the top level, there are probably only like five steps. Dig as deep as you want into any of those steps (i.e. engineering).

Tao has a spiel in his recent interview with Dwarkesh where he says that AIs are very good at explaining things - so just have the AI explain the proof in a human-friendly way.

I don’t see why it makes a meaningful difference if a human solves a math problem versus AI - it seems like the same amount of understanding will come out in the end. Either the understanding will come from humans arriving at the proof in the former case, or the understanding will come from humans understanding the proof that the AI came up with in the latter.

> To anticipate a common question: why couldn't my-unless be a function? Function arguments are evaluated eagerly, before the function ever sees them.

Interesting, so if you’re using a lazy language then you don’t need a macro here and could write my-unless as a function.

Location: Tampa, FL

Remote: In-person/hybrid/remote

Willing to relocate: Austin, TX

Technologies: Kibana, Looker, Python, Docker

Résumé/CV: https://docs.google.com/document/d/e/2PACX-1vQk-1zQa7PPg-3kq...

Email: edward.banner@gmail.com

Technical Support Specialist with 5+ years software engineering background. Diagnosing and resolving technical issues, communicating clearly with both technical and non-technical users. Background includes API debugging, cloud infrastructure (AWS), Python, and data systems.

Currently pursuing customer/technical support roles. Open to support engineering roles.

The music reminds me of Mappy Land (one of my favorite childhood games).

> You sound like a cigarette company rep.

> "I've seen an uptick in vague ominous attacks on Marlboro saying that our products kill people.

Then don’t vaguepost.

> Trust me when I say you don't want to live in a world where they win the AI race (if that wasn't already blatantly obvious).

They’re not perfect, but they’re the best organization in the running (with the second best person at the helm after Demis). The rest of the field isn’t close.

Nice FUD.

I’ve seen an uptick in these types of comment on HN - vague ominous sounding attacks on Anthropic.

It honestly seems like a coordinated PR effort from OpenAI folks to “both sides” the companies.

Which would, if true, be a further indictment of the culture at OpenAI.

I love the timer. It’s the closest game to ThruLine I’ve come across (that’s a compliment).

I’m an American and I just had the thought - if I was working in Japan at a Japanese company and I had the opportunity to hire, would I have a bias to hire other Americans?

Honestly probably, since I understand them the best.

Early on in my ChatGPT usage, one of my messages got interrupted/cut off (as happens occasionally).

My first thought was "oh they're going to need to add a UI feature to allow me to click and tell them to continue the conversation".

Then I realized I can just ask the model to continue, obviating the need for a button.

That was a pretty mind blowing moment.

I see.

Have you ever tried paredit? It’s pretty much a lifesaver for this kind of thing.

Ah I see. That makes sense.

That’s making me really thankful to be a paredit user.

> I do wish there were an easier way to move in the ]}]})))}-ness of block ends though.

I’m not quite sure what this means. How is it different/worse than all parens..?

fyi I use paredit and just hit ) and it moves me past any kind of paren/bracket. But even without that you can just hit left and right..?

I have fond memories of cs224d [1] taught by richardsocher. It’s a bit dated at this point as it was created in the pre-transformer era, but it was a very cool introduction to applying deep learning to nlp at the time.

[1] https://cs224d.stanford.edu

I’m thankful for Racket - it got me regularly programming in lisp by virtue of LeetCode accepting it as one of its languages.

I did start to feel Racket’s “wordiness” towards the end - it started to feel a bit like COBOL. I’ve since moved onto Clojure and really appreciate the shorter keywords/function names/fewer parenthesis.

I still miss for/fold though - that thing is an absolute machine.

They are published on the leetcode website. Here's a screenshot showing where to click [1].

LeetCode does have an "unofficial" API to get the problem of the day. Should work for previous days as well. This code worked last I tried [2].

[1] https://github.com/ebanner/daily-coding-challenges#leetcode

[2] https://github.com/ebanner/get-daily-leetcode-problem/blob/m...

Nice to meet someone else that does the daily leetcode! It is really a nice feature - I wish every online judge website would add it.

Bonus - Racket is an accepted language on leetcode ;-)

I had a little excursion into Dyalog APL recently and wound up writing an emacs mode to evaluate Dyalog APL [1]. It was a pretty nice experience using Claude to extract the small subset of features I wanted from gnu-apl-mode [2] to work with Dyalog APL.

I’d really like to properly get into APL though. My plan is to solve a bunch of problems on Kattis [3].

I'm really enjoying this way of learning a new language in the age of LLMs - starting with easy problems on an online code judge website and work with an LLM to come up with/explain simple solutions. It gives me dopamine hits, lots of reps, allows me to start coding right away, and is a nice way to slowly ramp up difficulty and get practice with different features of the language.

[1] https://github.com/ebanner/dyalog-mode

[2] https://github.com/lokedhs/gnu-apl-mode

[3] https://open.kattis.com

Love the story and the article. The only nit I have with it:

> “His answers are… understandable, and maybe in some ways more digestible than we would get from an expert,” he said.

This does not reflect his actual responses? The interviewer keys off his most emphatic sounding words to keep the conversation flowing, but his answers are generally inscrutable.

He did a great job given the cards he was dealt though.