binarymax

I think about where this is going to end up in 20 years and it terrifies me.

They are adding up. They can ignore them, and when they are out of office, the reckoning will come.

"Unlike traditional monitors that force your eyes to focus at a near distance, Phantom allows you to look through the display and focus on objects at varying distances. This helps reduce eye fatigue during long work sessions by giving your eyes natural opportunities to relax and refocus."

Is there any science behind this or is it just a "sounds about right" claim?

I'm absolutely not arguing that X is false, because I don't know what X is, and I am arguing in good faith. I will follow up with the question: if the non-profit and the for-profit are owned by the same shareholders, what is the theft? Is this not a legal transfer between business entities?

I am also a somewhat harsh critic of Sam Altman (mostly around theft of IP used to train models, and around his odd obsession with gathering biometrics of people). So I'm honestly looking for answers here to understand, again, what wrongdoing is being done?

I want to understand this more, so can someone please ELI5 what the theft in the article actually is? Theft implies someone lost something. I think it's theft from the non-profit? But what does that mean? Is it theft of taxes because of the wealth accumulated in the non-profit was not taxed according to how it would have been for a for-profit entity?

EDIT: I'm not sure why I'm being downvoted. I read the article and it's not clear to me. The entire article is written with the assumption that the reader knows what the author is thinking.

The mondAIs

This is it right here. My foil is the Elasticsearch replacement because PG has inverted indices. The ergonomics and tunability of these in PG are terrible compared to ES. Yes, it will search, but I wouldn’t want to be involved in constructing or maintaining that search.

Well, it’s 2025 and I still need to install a 3rd party toolbar calendar app, so there’s that.

I also can’t snap windows, and Cmd-tab still can’t tab between different windows of the same application.

There’s lots more usability that can be improved IMO

Because, assuming it's done properly w/o cheating, it's a great learning tool. It's sometimes easy to forget that certain tasks are the way they are because they're supposed to teach. We don't structure teaching and learning around what the least painful thing is.

My two cents about this after working with some teachers: this is a cat and mouse game and you're wasting your time trying to catch students writing essays on their own time.

It is better to pivot and not care about the actual content of the essay, but instead seek alternate strategies to encourage learning - such as an oral presentation or a quiz on the knowledge. In the laziest case, just only accept hand-written output - because even if it was generated at least they retained some knowledge by copying it.

> MSRC bounty team determined that M365 Copilot was out-of-scope for bounty and therefore not eligible for a reward.

What a shame. There’s probably LOTS of vulns in copilot. This just discourages researchers and responsible disclosure, likely leaving copilot very insecure in the long run.

I ran a test just now, and gpt-5 without streaming works without the biometric check.

GPT-5 works, just not GPT-5 streaming. I posted about this a little while ago with more details: https://news.ycombinator.com/item?id=44837367

You can spend the time to learn the odds, and play the odds. Most people don't have even that basic skill.

Python has decorators which can be used to add sugar to methods for things like true parallelization. For example, see modal.com’s Python snippets.

https://modal.com/docs/examples/batched_whisper

Hey I know picking names is hard, but this is the first time I've ever been on the other end of a collision like this.

I think the point is: why is OpenAI wasting its time on this? If it's just another channel for billing tokens then OK I guess, but it's not like it's a huge breakthrough.

OpenAI should be the roads, not the trucks. Let other product teams sort out the AI browsers. OpenAI has lots of problems to solve related to models and thats where they should focus. This is a side quest.

Interesting as I was in the (very large) camp that never considered it for generation, and saw it as a pure encoder for things like semantic similarity with an easy jump to classification, etc

Do we need language specific LLMs? I can’t vouch for the data coverage or accuracy of Arabic in the leading models today, but I do know them to be highly cross-lingual capable.

Not necessarily. You can get an interest rate lower than your investment return rate.

You know, I had spent a good amount of years not having even a single thought about rational rose, and now that’s all over.

The invasion of Ukraine by Russia is the result of Russian instigation.

I'm currently blazing through "Chip War" and can't put it down. This news is fascinating in that context. I highly recommend the book to anyone who hasn't read it.

Per month or year? And what region?

There is no accurate valuation in this space.

Anyone else have a .io domain and worried about the future?

Not true. Whole classes of bugs are just eliminated - most importantly NPEs. I’ve seen tiny scripts in node that crash due to lack of an exception handler, and I’ve witnessed countless runtime errors in Java in the simplest of web services.

The way I think about rust is this: you can experience the pain up front while you’re getting your program to compile, or you can experience pain later during runtime in random unpredictable ways. Rust is the former. It takes time to learn how to reduce the up front pain, but I’m so much more confident in the end result.

I say this having years of coding background in languages like C, C#, Java, JavaScript, Python, etc.

What do you mean by this?

Best of luck and wish I had something like this a while back. Took me 4 years to figure out my triggers and one of them was pure luck.