I never saw this or any ad on YT, using only the Vivaldi trackerblocker and uBO Lite, also the Vivaldi adblocker works fine, but is discovered sometimes by the antii-adblock algorritm from YT. Anyway not possible to avoid the promotions from the Youtuber itself (not confuse with embedded sponsorblocks, which are filtred), also not posible to watch vids "only for members" which are appearing more and more, but I think it's a legit manner to create incommings for the authors.

Chinese people are heavy controled by the gov, but they have more civil rights, health care and social services than the US never had in their history.

ChatGPT - Apertus (FOSS) (Swiss)

OneDrive - Filen (Germany - 10GB free)

Google Photos - vgy.me (UK)

Gdocs/Gdrive/Gmail - Murena Workspace

Google Maps - CoMaps (FOSS- privacy focused)

NSFW isn't only porn. Use Swisscows as search engine and you'll understand why it isn't recommend to block NSFW if you aren't a young kid.

No problem

Agree with it, despite using an AI search since almost 4 years, Andisearch, because it's the only one I tested with a pretty good accuracy (~90%). Anyway, it's always needed to contrast the information in the web, independent of results from AI or not, a lot of BS out there.

To this topic Andi said:

Based on recent examples from 2025-2026, AI search engines frequently provide confident but incorrect answers, demonstrating several key problems:

1. Inconsistent results - Siri with Apple Intelligence gives different wrong answers to the same question when asked multiple times[^8].

2. False confidence - AI provides detailed but completely incorrect information, like Google AI claiming a South Dakota team won North Dakota's championship[^8].

3. Regression in quality - Traditional search results often work better than AI versions. As John Gruber notes, "old Siri... at least recognizes that Siri itself doesn't know the answer and provides a genuinely helpful response"[^8].

4. Poor accuracy even on popular topics - Siri achieved only a 34% accuracy rate when asked about Super Bowl winners, with one stretch of 15 wrong answers in a row[^8].

The core issue appears to be that AI search engines prioritize providing definitive-sounding answers over accuracy, making them less reliable than traditional search results that simply link to authoritative sources.

[^8]: Daring Fireball - Siri Is Super Dumb and Getting Dumber

Ads? Ah, those annoying stupid things I watch in TV. Online I don't see any of these since more than 10 years.

I never download apps from other than the official homepage or from sites linked from these. This case is by download 7-Zip from a fake homepage, not from the real one. This can happen when you use a third party download page, even if Ninite is normally a trustworth site.

Enough security and privacy with Portmaster, no need of any AI crap in the system.

Use one from Musk

It won't

Normal - 4h coding + 2 h debugging

With an LLM - 5 min coding + 24h debugging

There are tons of pages out there with lists of code snippets which can help the devs without the need to use an LLM. It's naiv to think that LLM which hallucinate in a normal answer won't do it with coding.

The best weather forecast is always the one from the official weather agency of your country