timhigins

Acid3 is 100 tests but to have a browser that works on any real percentage of the internet you need to pass the over 52000 tests and nearly two million subtests [1] that make up the Web Platform Tests [2]

This is like the difference between a college degree and passing a spelling test.

1. https://servo.org/blog/2023/07/20/servo-web-platform-tests/ 2. https://github.com/web-platform-tests/wpt

Not to mention this site is clearly vibe coded and has a vague call to action and no source code available thus indicating possible vaporware.

most people I know who have worked at Amazon hated it because of how limited of a scope each team has and the extremely slow pace to get anything approved or deployed. You may have different luck depending on what team you’re on.

I recommend you join a start up, specifically in reinforcement learning as a service, human data, or a neo lab. This will give you direct visibility into the current frontier tech and allow you to learn and grow the most because there will always be things to do.

I recommend https://github.com/wesm/agentsview - its quite good, has usage metrics and nice viewer/search for transcripts from most agents.

As noted above, I'm not the creator of Paseo, just a big fan. Beautiful is just ways to describe it (along with convenient, and powerful) since it has a very focused and clean UI. Especially compared to many open source projects, which often don't put in that much effort or are unabashedly vibe-coded.

Of course it is open source so I hope some of the designers/people who've commented on this post can maybe contribute ideas to improve it even more. I have noticed a few places where some common actions take 1 or 2 more clicks or taps than they should - things can always be a bit more convenient and beautiful.

But overall I'm incredibly impressed and you can see some examples of the focus on simplicity and a nice UI, and follow the creator here: https://x.com/moboudra

what is the goal of this software? how close is it to success?

measuring lines of code is meaningless without its real world or business impact. apologies if this is described in the docs, went through the repo but couldn’t tease that out

Yeah I saw some issues this week on our Go service

This kind of reads like an action or war novel

It would be really nice to have a tag on HN to filter out LLM-generated, or at least partly AI-generated content like this.

If an article makes it to the front page despite being AI-generated it probably has some interesting points or ideas, but it's unfortunate that people seeem to choose the speed and style of LLM writing over the individual style and choice that made the writing of yesteryear more interesting and unique.

Darn, more LLM-generated blog posts?

This was published by an anti-vaxxer/vaccine denier and also seems to be AI-generated. Would recommend linking to the original study instead. The homepage of the site includes articles like "Do Viruses Exist?" "(POLL) 96% support federal control of DC to fight crime" "Autism Spectrum Disorders: Is Immunoexcitotoxicity the Link to the Vaccine Adjuvants? The Evidence" and so does his twitter page: https://x.com/NicHulscher.

Looks great! Diving into the docs I especially liked the idea of a headless React library so I can design my own UI and add some extra components. How difficult would it be to automatically highlight or underline certain terms in the PDF and then render a custom component when I click or hover over the term?

I opened up the developer playground and the model selection dropdown showed GPT-5 and then it disappeared. Also I don't see it in ChatGPT Pro. What's up?

How well does this work with partial downloading of a file based on column statistics via HTTP Range requests?

Any plans to integrate FastLanes into DuckDB? You would instantly get more usage from its SDKs in many languages and the CLI.

Any comparisons with Lance/LanceDB?

good points, but partly ai generated?

Exactly and see also ROCM/HIP which is AMD’s reimplementation of CUDA for their gpus.

You can test it out at https://opencut.app/projects. I added some screenshots to the README here (https://github.com/alexkreidler/OpenCut/tree/main?tab=readme...) and opened a PR to merge the update upstream.

It’s known for deep tech/hard tech companies, especially in space, aviation, and energy. Many are in El Segundo or were founded by SpaceX/Tesla alumni.

https://www.addtheegg.com/p/2024-la-hard-tech-50

https://www.vanityfair.com/news/story/el-segundo-california-...

https://bowoftheseus.substack.com/p/south-bay-la-is-ground-z...

what neighborhood did you choose? was it santa monica, venice or mahattan beach? did you have any second choices/other favorites?

Great to hear, thanks for the response!

> we're open sourcing everything

Does/will this include the training data, hyperparams, and weights for the models?

People will be reluctant to buy an "open source" robot if the key ML magic to make it work is closed off, e.g. if you charge a subscription for it.

Yeah this would be interesting. Also the founders should consider open sourcing some/all of the code. It could be an awesome platform for the open finance community.

If you can prove the data was altered or tampered with you'd have a solid fraud case though I imagine proving it is the difficult part.

Based on satellite imagery from Maxar and reports from Iran state news agencies, they moved the uranium in about 50 trucks to a “safe location” [1].

From Reuters [2]: > Hassan Abedini, deputy political head of Iran's state broadcaster, said Iran had evacuated the three sites some time ago. > "The enriched uranium reserves had been transferred from the nuclear centres and there are no materials left there that, if targeted, would cause radiation and be harmful to our compatriots," he told the channel.

1. https://www.newsweek.com/iran-nuclear-strikes-us-donald-trum... 2. https://www.reuters.com/world/middle-east/iran-israel-launch...

The title of this post really doesn’t match the core message/thesis, which is a disappointing trend in many recent articles.

> unilateral international regulation

is an oxymoron/contradictory

https://huggingface.co/datasets/institutional/institutional-... https://huggingface.co/datasets/institutional/institutional-...

https://github.com/instdin/institutional-books-1-pipeline https://www.institutionaldatainitiative.org/institutional-bo...

Note that Trump's Big Beautiful Bill as it passed the House of Reps would bring back 100% expensing of R&D expenses including software development costs/salaries.

> Exempted Code

> Not all source code, documentation and metadata used in the development of Direct File is included in this repository. Specifically, any code or data that is considered Personally Identifiable Information (PII), Federal Tax Information (FTI), Sensitive But Unclassified (SBU), or source code developed for National Security Systems (NSS), as defined in 40 U.S.C. § 11103, is exempt. Due to these restrictions, certain pieces of functionality have been removed or rewritten.

Very curious about what these pieces are that were removed

The code that defines how the fact graph works is here: https://github.com/IRS-Public/direct-file/tree/main/direct-f...

but the actual tax definitions that deal with facts and derived calculation are here: https://github.com/IRS-Public/direct-file/tree/main/direct-f...

See for example the standard deduction and tax rate calculations https://github.com/IRS-Public/direct-file/blob/main/direct-f... https://github.com/IRS-Public/direct-file/blob/main/direct-f...

I imagine these are based on the MeF (Modernized e-File) schemas because the system needs to transform the input data into XML MeF schemas to submit electronically to the MeF system (See https://www.irs.gov/e-file-providers/modernized-e-file-mef-s...)