AI scrapers use these residential proxies. It's not (just) VPNs and Tor routing. Several bottom-feeding companies openly advertise such scraping services, for pretty much any country you may want. I administer a wiki that's been on the receiving end of such scraping, and the majority of these scraping requests are in fact coming from residential IP-addresses rather than data centers.

I don't know whether these are hacked accounts, people getting tricked or paid to run these scraping apps on their devices, but it's impossible to block them all. Even if you let fail2ban block entire /24s for every detected hit (even disregarding the collateral damage and the fact that these blocks don't solve the issue, the fail2ban and iptables overhead starts to outweigh the apache load at some point).

Anubis seems to be taking care of it for now, but it's obviously only a matter of time before they can deal with that one too. Although its delay does enable fail2ban rules to block the IP-addresses before they get to stress the mediawiki php scripts, attempting to diff 2 revisions of a random page from 10 years ago.

For the Irish language course the recordings of native speakers were taken offline in 2023. The AI replacements are nonsensical.

This story is about AI generated courses, not voices, but my post was still (accidentally) on-topic: when they previously used AI to increase volume of content, they were ok with quality being thrown out the window.

The AI generated courses might be low quality, and the original (English) courses might also go downhill because the type of exercises they produce may now be restricted to the type of things that their AI is able to reorganise for other languages. E.g. it might go further in the direction of vocabulary memorisation.

Isn't it that they call it "español" in Latin America, and "castellano" in Spain?

Unfortunately, the voices are really bad.

It's a pity they don't also make available the old courses, with audio from native speakers.

Maybe from NOAA? (at least until they get defunded)

Nope. That's why I changed all my players to BlueOS.

FSF's Zoe Kooyman and Krzysztof Siewicz will give a presentation on Sunday 2nd of Feb:

"FSF's criteria for free machine learning applications"

https://fosdem.org/2025/schedu...

It'll be streamed. Well worth tuning in for. A recording should be online soon after.

ELIZA's evil twin.

I replaced all my SONOS connects with BlueSound node Nano devices. A pricey replacement, but worth it.

As a bonus I was now able to turn off SMB1 on my home Samba server !

> Every large NAS vendor (Synology, QNAP, etc) has their own SMB server they wrote themserlves

That's untrue. Both Synology and QNAP use Samba. QNAP contributes code and bugfixes back to samba.org (Hi Jones !).

It's indeed not AI, but it's not particularly related to quantitative easing either. It's a regular cycle that's been repeated since forever by publicly traded companies, and game companies are just getting better at playing that game as well (pun intended): https://www.youtube.com/watch?...

"Made in China" and then transported to and consumed everywhere else in the world isn't powered by fairy dust and unicorn smiles. It's easy to have low emissions when you externalise production. Let's see how it goes once (if) the process of bringing manufacturing back gets under steam.

The upstream Linux kernel doesn't differentiate between security bugs and "normal" bug fixes. So the new kernel.org CNA just assigns CVE's to all fixes. They don't score them.

Look at the numbers from the whitepaper:

"In March 2024 there were 270 new CVEs created for the stable Linux kernel. So far in April 2024 there are 342 new CVEs:"

Yes ! That's exactly the point. Trying to curate and select patches for a "frozen" kernel fails due to the firehose of fixes going in upstream.

And in the kernel many of these could be security bugs. No one is doing evaluation on that, there are simply too many fixes in such a complex code base to check.