rewq4321

Not just the API - I'm now suddenly having DNS troubles with one of my domains. I think maybe only domains/sites that use their "configuration rules", or some specific feature like that.

> Also the attention mechanism is baked in during pretraining

IIUC, this is no longer necessarily true with positional encodings like ALiBi: https://github.com/ofirpress/attention_with_linear_biases

What source is that understanding based on? I was not under the impression that Anthropic was intending to open source their models at all.

Yep, unions are fighting against even allow moving to 1 crew member per train, let alone zero.

https://smart-union.org/our-priorities/legislative-issues/

https://public-inspection.federalregister.gov/2022-15540.pdf

*make their neighbourhood more accessible to those who are not well-off

Are you sure you're up to date on the latest additions to this API? This wasn't in the MVP:

> The origin private file system provides optional access to a special kind of file that is highly optimized for performance, for example, by offering in-place and exclusive write access to a file's content.

https://web.dev/file-system-access/#accessing-files-optimize...

It was originally going to be a separate high-perf "Storage Foundation" API, but that was merged with the File System Access API.

The root cause is primarily housing policy. Let developers build many apartments, and let them build small apartments. Zoning laws that make it a years-long process to just get a permit, or that force developers to build apartments with minimum square footage, or prevent large families living in a "small" apartment - these are all luxury laws forced upon poor people so that rich people can be more comfortable.

I count myself as progressive but am almost disgusted by how many progressives try to muddy the water here because they don't like to face the fact that they have to give something up to make their neighbourhood more accessible to those who are well-off people.

Anyone who doesn't take housing policy seriously on the issue of homelessness is either naive, or doesn't actually care about the issue and instead is trying to use it to push some other agenda.

Fun fact: It's also far more environmentally friendly in terms of how much land is required to produce it: https://ourworldindata.org/palm-oil

Not the person you're replying to, but if you're willing to do that to your walls, you're probably willing to pay for an external antenna and internal transmitter

> there is value in publicizing the fact [...] This has consequences in how they approach treatment (medication-centric)

If anything (in my experience) people are too biased against medicine. And as others pointed out, this isn't at all evidence that medicines don't work - their efficacy is based on double-blind studies, not on chemical imbalance hypotheses.

There are many other possible mechanisms that aren't related to the environment. Publicizing studies like this beyond academia is only likely to cause confusion and unfounded anti-medicine sentiment.

As of quite recently that may be changing in Chromium (early experiment): https://groups.google.com/a/chromium.org/g/blink-dev/c/9xWJK...

IIUC, the TLDR is that they're experimenting with loading a bunch of popular scripts (e.g. top `N` scripts from CDNs) on all browsers, so the response is immediate, but they can't use the timing for fingerprinting because it's immediate for everyone.

The holy grail for me would be a tool that allows me to run (basically) any Python package in the browser.

Pyodide is at the point where it can run any pure Python wheel out of the box, but building non-pure wheels (e.g. containing C, C++, or Rus) is a very manual process.

I don't think we're tooo far away from having tools that automate all the manual work, but I really wish we had them today.

That doesn't make sense as a typo? Node.js runs on V8 too. And whether two platforms run on V8 doesn't say much about their compatibility outside of "they both run javascript".

Those interested in this should check out Pyodide[0]. It basically "just works" so long as the libraries you import are pure Python or are part of the core scientific stack (the Pyodide authors have done the work to manually port all the C code behind numpy, scipy, etc.).

What I really wish for is for ~all Python packages to work in the browser without manual porting of the underlying C/Rust/etc. being needed, since a lot of the interesting and useful libraries aren't pure Python, and manual porting is non-trivial.

I'm not sure what the best route to that future is, but I'm guessing it'd probably help if Python had a wasm runtime in its standard library[1], since then authors of libraries that use C/Rust/etc. might make cross-platform builds (perhaps by default).

Regarding this Pycon speech, it seems that it's related to the following entry in the 3.11 changelog[2], which the speaker was heavily involved with:

> CPython now has experimental support for cross compiling to WebAssembly platform wasm32-emscripten. The effort is inspired by previous work like Pyodide. (Contributed by Christian Heimes and Ethan Smith in bpo-40280[3])

But maybe Christian has more to reveal here? In any case, I'm hugely appreciative of all the work that is being done to bring Python to the browser!

[0] https://github.com/pyodide/pyodide

[1] https://discuss.python.org/t/add-a-webassembly-wasm-runtime/...

[2] https://docs.python.org/3.11/whatsnew/3.11.html

[3] https://bugs.python.org/issue40280

Why does it need to be direct? A bit of glue code doesn't hurt, especially since most of the other glue is going away with WASI.

Not infinity, but yeah it's worth more than people generally think. But in the end you don't really lose many clock cycles anyway because everything actually runs in C/CUDA/etc. behind the scenes

Why? If your messaging app doesn't have the ability to block stranger requests, why are you still using it?

And even if you for some reason don't want to restrict your requests, you'll probably still be fine - Gmail handles protects me from spam pretty well.

+1 and you can even connect your Colab to a GCP Marketplace Colab runtime that has no time limit (but will cost us) if you e.g. need to run something for a few days (although then you don't get the awesome Google drive mounting - hope they fix this eventually)

Users care about UX. Ads affect UX. If there is healthy competition, then things can't get as out of control as you're implying.

Ads aren't a problem if there are multiple competing services to choose from. If customers care about the ad-related UX then they'll switch to a better competitor. The problem is monopoly (as usual).

I recently switched to a new tech publication because the old one had a "page jumping" problem (related to ads) that they didn't fix for a while even after I sent them a message. No big deal - this is how a good market works.

Another problem is bad discovery mechanisms. Google search should de-rank pages with to many spammy ads, but that's not in their interests due to their control of the ads ecosystem. This does kind of come back to being a monopoly problem.

A cool feature of ads that's not mentioned in the blog post is that they automatically price discriminate. Rich people effectively pay more and subsidise the content for those from developing countries. This probably angers rich people more because they (somewhat subconsciously) value their attention at a rate which makes the content seem expensive.

Of course, there are still problems to solve with malware, excessive tracking, etc. but these don't seem intractable.

You're assuming there will never be competition in his niche or that he will somehow discover a way to maintain a monopoly.

This seems very impractical given the way the internet currently works. Most startups use dozens of SaaS products, let alone more basic/foundational things like global CDNs. You're being logged at every step of the process if only to prevent spam/DDoS/etc.

What you're asking for would require a fundamental restructuring of the internet, and of software business models, and a lot of other stuff. I can't see that happening any time soon.

In the meantime you can try using Tor, but good luck not getting blocked on half the websites you want to visit - and you can't blame the website for that (they need DDoS/spam defence).

Almost comical to draw an analogy to 3D glasses imo - VR is a completely different ball game. VR headset purchases are growing at more than 2x year on year, and the "try it once and forget about it" thing seems to be less and less of a thing based on post-Christmas 2020 & 2021 WebXR usage stats that I've seen.

The virtuous cycle is starting now that big AAA devs are taking notice of the platform. More games => more users => more devs => more games.

That's not at all my experience, or that of my friends. We play multiplayer games together via streaming (Geforce Now) even though we're in different continents, and it works perfectly fine.

Exactly. So many comments here are acting as if they're forced to use npm+react+tailwind+etc. but you're absolutely not. Web standards (querySelector, fetch, display:flex, ...) are so far along compared to a decade ago that you could use zero libraries/frameworks and be significantly more produtive than we were back then.

Web standards still have a long way to go, but it's comical to claim that things have gotten harder.

It's easily an impressive enough leap to get hundreds of upvotes. It's about that fact that it takes two clicks (including waiting for arxiv's absurdly slow PDF download speeds), and then some scrolling to get the results.

As someone else said, the blog post that they'll likely release will hit front page.

> 0.01% of them got to have some cool hallucinations though - look on the bright side!

Hmm

The food/diet component of their existence is really the absolute least of their problems.

The video by CPG Grey is better, as someone else already mentioned: https://www.youtube.com/watch?v=cZYNADOHhVY

I think there's a general failure of communication here, and throughout the current social discourse on "cynicism".

There's a sort of false trichotomy at play in this post, and on Twitter I've noticed a dichotomous version: "Should we be happy that we're improving, or sad that there's so far to go?" (That's my summary of it, anyway)

Why not both? We should appreciate how far things have come (it is truely astounding), but of course that doesn't imply that we should be content with where we are. We should be deeply discontent - at least to the extent that it motivates us to fight towards a better future. A magnitude or type of discontent that leads to depression/apathy is obviously not helpful.

It seems like a significant portion of tribal "controversy" on social media is stoked by mouthpieces on either side that are almost purposefully playing into the dichotomy perspective, because adding subtlety into the discussion would throw water on the flames, and that's just ending the "fun".