throwaway89201

In the short term I'm sure the approach works, but the proof of work part of it is just theatrics and doesn't stand up to scrutiny. Using an uncommon CAPTCHA with some alternative heuristics is enough to deter spammers (until it isn't).

Any back of the napkin calculation will show you that if spammers actually start solving the challenge, they will either pay a ridiculously low amount of server CPU time, or regular users on relatively underpowered devices (i.e. smartphones) will have to wait forever for a pageload.

You'll never see that basic calculation being spelled out on product pages, because it will show its mostly smoke and mirrors.

This website seems almost dangerously wrong. It uses only a single phoneme / sound for single letters, regardless of the actual phoneme used in the word. For example, every 'a' in a word is the /æ/ phoneme (as in "cat"), but this same sound is used for the word "a", which should be pronounced as /ə/ or /eɪ/ (as in "about" or "cake").

It also seems to mix up American and British English pronunciation between the phoneme and word forms a lot, for example in the word "class" which uses American phonemes but uses the British word pronunciation.

It sometimes cheats by declaring a word 'difficult'. For example word "no" is declared difficult, which the average learner is probably pretty practiced in saying. It's just not using the same phoneme as the all the other 'o' phonemes on the website.

I'm sure its made with best intentions, but the foundation on which the app is built / prompted is incorrect. Please don't use this to teach children. Use proper material made by educators, acclaimed authors and experts.

> but not the beginning

You didn't stop reading after "The hack itself is on its way out, but the design behind it still matters for your code"? I did.

That xkcd is exactly what this thread is about, and so your lazy linkdrop is completely beside the point. The bots are currently posting lots and lots of (often superficially) "constructive and helpful" comments, and it turns to be a complete slog and burnout for humans to figure out whether to press the "constructive" or "not constructive" button.

Thanks for that perspective! I think that dynamic is present in quite a lot of other communities too, especially the ones who have not traditionally embraced open source. You also see it a lot around creative / artsy work where it almost seems to be a faux pas to share anything but the final product. People do recreate things or just are inspired by other work. Coming from open source software it always seemed like wasted potential to me, but that's probably just my bias.

> In a sense, these mods are no longer open source, but rather source available

I understand what sense you mean, but the terms "open source" and "source available" have very clear definitions, and it has nothing to do with accepting contributions or not, only with the license the code is published under. So in a sense only about downstream effects; upstream effects are still a courtesy also with open source.

I find Photopea to be completely legitimate, and ripoff is too strong, but you can recognize that it's not just a "fairly feature-comparable alternative". It has most of the functionality at the exact same place, with the exact same shapes, because that will help user if they already know how to work with Photoshop. That makes Photopea a well-executed clone, one that has been reimplemented on another platform from scratch.

Mistral STT offers it, which is an open model.

The device runs Android, which makes it very simple to use the Keystore system and to store a device specific private key within the TEE or SE where it can't be very easily extracted. If you really don't want to provision in the factory, you could use secure boot measurements to do it remotely. Of course this isn't completely watertight either against a physical attacker, but it would survive a filesystem dump attack and is the least you can do to appear competent.

They can only do that through an "SSL added and removed here ;-)"-like 'cooperation' and not through passive listening because with forward secrecy a certificate key doesn't secure the transport encryption directly. They could actively MITM outside the perimeter of the target by issuing a new certificate, but that would show up on CT logs.

I'm really hoping Mistral will succeed with their open models, so I'm a bit biased, but I don't have any affiliation. This submit is a bit of well-meaning but confused scaremongering however.

Mistral has had, and continues to have, a toggle in the admin settings that permanently disables training on your data. The option has not been removed, and previous opt-outs are still honored. As far as I know, Mistral always trained on your data by default except for the enterprise plan, with the option to disable it on all plans, and with the option for organizations to make the choice for all your users.

Kagi Ultimate still uses mostly closed models by OpenAI / Anthropic, etc.