Hacker Newsnew | past | comments | ask | show | jobs | submit | plastic041's commentslogin

It's really frustrating that Dario acts as if he's not the CEO of one of the world's most advanced AI companies. He can just slow down his own company. Of course he doesn't want that. He wants to slow down other companies, but not his own.

Also, regarding the incidents: Neither he nor Sam Altman takes responsibility for those incidents. You can't say, "Wow, someone's agent is gone rogue; let's slow down" when you are literally the person in charge. CEOs and researchers will only slow down when they realize that they will face consequences if their LLMs misbehave.


To be fair, this is a classic prisoner's dilemma. If the others don't play along then there is no benefit at all, it's all negative for Anthropic.

> Neither he nor Sam Altman takes responsibility for those incidents.

He doesn't need to take responsibility - as CEO, he has that implicitly. And we know it.


He can slow down his own company kind of like how Zelenskyy can just declare peace in Ukraine. It works a lot better if you can get the other sides to agree.

Give that Dario is one of the frontiers of LLM development, literally started the LLM race, and have been dominating the market, so he'd be Russia, if we have to use the war analogy.

He took every benefits of being frontiers and now he's kicking the ladder.


It's frustrating to see smart people failing to recognize a prisoner's dilemma when they see one.

What is the prisoner's dilemma here? I am a dumb person

"I'm not going to let someone else take the credit for ending human civilisation!"

Or like Von Neumann's (paraphrased) "you need to confess a crime to take credit from it"

So in this example, the equivalent of Zelenskyy and the Ukrainian people fighting for their lives and the very existence of their country for Dario is... losing lots of money?

What a ridiculous analogy to make.


If you played out the hypothetical that they're earnest and don't care about the money, then what?

I can't stop thinking that these use cases aren't real, and OpenAI is just gaslighting people. Humans don't need AI for these trivial things.

This ad feels like a science fiction, where humans can't do anything without the help of their AI assistants.

- Design a tattoo of my cat

- What hairstyle should I have

- Arrange these flowers

What next? Who should I date? Open the door?


You're right that no one needs this, but OpenAI are right that many people will use it

Former discussion (85 comments):

https://news.ycombinator.com/item?id=47434732


You can read these two articles instead of an article written solely by a LLM.

- https://www.mathpuzzle.com/MAA/41-Sudoku%20Variations/mathga...

- https://www.nikoli.co.jp/en/puzzles/sudoku/


Thank you!


All examples run terribly, like 20fps on m4 pro. Also, examples:

- Liquid metal. I don't understand what this is but this is definately _not_ how 'liquid metal' would look like. Also stickers glitches on edge.

- watercolor. Looks like its blending uses RGB?

- ASCII. not even ASCII. Also why does this run terribly? It's just a simple 3d model with a shader, it should run 1000+fps.

- Bricks. This looks fine. not impressive though.

- Shaders. Refraction shader.

I don't trust someone who says their tool can generate ASCII art renderer without knowing what even ASCII is.


I thought too long to find out what a means in a new runtime.


Sorry about that. Basically you can now write and run code in the K/q/or qsql languages on your laptop for free. For a very long time (20 years+) this language has not been accessible to most - it’s almost exclusively used on Wall Street and has a very high price barrier to entry. Q programmers are still some of the most highly paid engineers consistently - as the workforce is small and the use cases are (generally) extremely lucrative. If you are a programmer wondering what comes next… q might be for you. Over the decades I’ve taught a few dozen people Q- and I’m happy to say they’ve gone on to have wonderful and stable careers in high paying jobs with deep stability. Mostly hedge funds and banks (and exchanges). It’s a wonderful and often times world changing language and now you can play with it yourself. It’s also very very fast. As a database it’s still insanely great for many usecases.


Would love to talk. I sent an email to the maybe-address string in you profile, but it bounced.


Categorizing Japanese verbs as -ru or -u requires more context.

I prefer the term "group 2 verbs" to "-ru verbs." Group 2 verbs are verbs that end in -eru or -iru, not just -ru. Of course there are some exceptions, like kaeru, which ends in -eru but is actually a -u verb. Conjugation is easy: remove the final -ru and append -masu, -mashita, etc.

"Group 1 verbs" (again, -u verbs) are verbs that are not group 2 verbs. Conjugation is a bit more difficult because the -nu, -bu, -mu, and -u verbs have many suffixes. However, after memorizing these two (-nbmu and -u, because -nu, -bu, and -mu are almost the same), the rest are easy.

There are only two irregular verbs: kuru and suru. Just memorize them.

I learned Japanese by just memorizing. Once you have memorized enough verbs and their conjugations, you can figure out the conjugation of a new verb even if you don't understand how it works.


there are more irregular verbs than just kuru and suru. iku and aru are also irregular, for example.

Irregulars notwithstanding, the conjugation pattern is actually completely lossless if you just remember the imperative form (e.g. 着ろ kiro, 切れ kire) instead of the infinitive, which is lossy (e.g. 着る kiru, 切る kiru). Then there's no need to have to remember, "oh... is this -iru verb group 1 or group 2?"


They’re sometimes called “semi-irregular” because they are mostly regular with, like, one deviation. The list is not long and it is quick to memorize.


Iku is kinda irregular but it's only itte and itta instead of ite/ita. Also aru/nai is more like antonyms rather than conjugation?


I wanted to check if the information on this page was correct. I started searching and found this site [0], which looks very similar. I thought it was made by the same person, but it's not. It's just another website designed using LLM.

One advantage of LLM is that you can quickly and easily generate a "pretty decent" website. However, there is a drawback, that there is a high chance that a page with a very similar design(and similar idea) already exists somewhere.

[0]: https://chinesecoloratlas.com/


I don't think it is a case that llm happens to create similar stuff. Either both are made by the same person or one is a copycat of the other.

What are the odds of both websites showing the same color today (Cochineal and 胭脂)?

And both use Roman numerals MMXXVI ? It makes no sense to use that in a "Chinese" color website.


I think they are made by different people/organizations. Chinese colors is by TOTM Tech, which sells various (pretty obviously vibe coded) design tools and assets. OP doesn't sell anything so I don't think they are related.

> MMXXVI

That too. LLMs love to writing year in roman numeral. https://github.com/search?q=%22%C2%B7+MMXXVI%22&type=code There are 2.8k of these in github.

> the same color today

Looks liks Chinese colors always show that color in Hero section? It's coincidence I think.


People should know that most of the Mac examples shown in the article are regressions that happened in Tahoe.

- Sequoia's Notes had perfect sliding animation.

- Safari's URL bar seemed the same. (Though I think it's more a placement issue, not animation issue. They could just put a placeholder at the end of the bar, not at the center.)

- 'Save as' dialog had a problems with sidebar, but not janky as much as Tahoe's.

- Zooming in preview app worked fine.


Looks clean and works fine, but it needs optimization. Clicking "Type" in the "schema builder" example takes 1~2 seconds to open the popover in the landing page(macbook pro m4). I think its because there are lots of heavy components, but still it's too slow.


it should be much faster on the individual component's page

the root page is a bit slow with all the viewers, in practice you probably wont have that many in your app on one page


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: