@SuperDicq >Can someone explain to me why one is wasteful and the other is not? I personally find both usecases to be legitimate actually.
The former actually has a practical usage - it secures such cryptocurrencies and allows transactions to be done securely.

The latter has no practical usage - it just wastes CPU cycles for the sake of wasting them and it does not stop scrapers, as they have more processing power to spare than those who run only free software and therefore is not legitimate.

>Personally I am not a fan of Anubis,
I am a fan of Anubis; https://www.gnu.org/software/anubis/ that is.

>it requires me to have Javascript enabled for your website. I usually don't want to enable Javascript on websites that I don't trust.
I just close the page if I see the that proprietary malware JavaScript is required (even if it under a free software license, the way it is served denies the users freedom and therefore is proprietary software for everyone but the web host).

>I understand why a proof of challenge might be necessary to prevent denial of service
PoW cannot stop big enough denial of service attacks unless the clients must connect to some other big server pool first and then can only access the server after doing the processing.

When it comes to LLM scrapers, you can counter them fine without serving proprietary software.

With minimal work and minimal processing power, you can;
- Setup bait links (several follows and you get IP nullrouted for a time)
- Setup ratelimits.
- Setup bait files that serve gzip bombs as "content type" text/html & "content encoding" "gzip".
- Setup a tor middle and see that the CCP goes and blocks packets from IPv4 Chinese scrapers with their firewall without you having to block anything (just make sure the site is also available over IPv6, so the Chinese can access it).

With more work and processing power you can serve delicious Marklov slop and poison the training data.