2mo ago

Automatically generated page summaries as an experimental feature of Firefox 139 nightly

www.omgubuntu.co.uk

Try Firefox's Experimental Link Previews with AI Summary

Latest nightly builds of Firefox 139 include an experimental web link preview feature which shows (among other things) an AI-generated summary of what that page is purportedly about before you visit it, saving you time, a click, or the need to ‘hear’ a real human voice.

30 comments

Hate having to read an article to understand what it's saying and would rather read what an AI says it (potentially) says instead?
This reads like satire.
- Maybe because it is, an article says later "Saltiness aside" 😏
  
  Oh! That makes more sense lol
- this is the next step from just reading headlines and getting everything wrong

- Ideally it would do that for links. As in, hover and it gives like a 2-sentence summary. In your example, the summary could say something like:
  Cashews are pretty good for you, but almonds are better. The article discusses micronutrient differences between the two.
  Or something like that.

Hell yeah! My local news sites always go on and on about stupid stuff before getting to the point. Which is almost always "we don't know".

If it works around paywalls, it could actually be a useful feature.
- It reportedly works entirely on your machine (as it meant to be privacy preserving by default). So it will probably see only the data you can see.

If there’s one thing I’m afraid of with Firefox is that recent AI bubble will lead to way too much bloat focus.

Awesome! Now half the sites on the web contain a tiny drop of information, buried under paragraphs and paragraphs of GPT-generated text, and now my browser uses the same LLM to reverse engineer the original information. What could possibly go wrong the LLMs talk to other LLMs and summarize, bloat, summarize and then bloat and then finally summarize every bit of information for us.
Do we actually still make websites for humans, or is it all just AIs and SEO?

This feels like windows recall...
- Except without the shitty parts where it keeps a log of everything you do and sends it off your device, luckily.
  
  Well, it is sending it off your device to the AI's API. Luckily it won't have any id information, such as cookies, screen size, OS, IP, etc.
  The problem seems to be with the word luckily.

I wonder if the preview does a pre-fetch which can be identified as such? As in, I wonder if I'd be able to serve garbage for the AI summarizer, but the regular content to normal views. Guess I'll have to check!
Update: It looks like it sends an X-Firefox-Ai: 1 header. Cool. I can catch that, and deal with it.
- I agree with your sentiment. It's sad that your (or my) website can't respond with a X-This-Website-is-not-AI-Garbage header to indicate that "Hey user, you can actually just open this website and read it and get the infos you need without an AI assistant.".
  I'm pretty sure Firefox's preflight request will also not load ads and so could be seen as a bad scraper.
- Definitely won't be visiting your website then if you intentionally fuck with people to make their browsing experience worse. I hate web hosters who are against the free and open internet.
  
  Pray tell, how am I making anyone's browsing experience worse? I disallow LLM scrapers and AI agents. Human visitors are welcome. You can visit any of my sites with Firefox, even 139 Nightly, and it will Just Work Fine™. It will show garbage if you try to use an AI summary, but AI summaries are garbage anyway, so nothing of value is lost there.
  I'm all for a free and open internet, as long as my visitors act respectfully, and don't try to DDoS me from a thousand IP addresses, trying to train on my work, without respecting the license. The LLM scrapers and AI agents do not respect my work, nor its license, so they get a nice dose of garbage. Coincidentally, this greatly reduces the load on my backend, so legit visitors can actually access what they seek. Banning LLM scrapers & AI bots improves the experience of my legit visitors, because my backend doesn't crumble under the load.
- User agents that /look/ legit, but the chances of someone using them /today/ is highly unlikely.
  I hope I'll never need to deal with websites administered by you. This is way overboard.
  Sajnos az esélye nem olyan kicsi mint azt szeretném.
  
  Overboard? Because I disallow AI summaries?
  Or are you referring to my "try to detect sketchy user agents" ruleset? Because that had two false positives in the past two months, yet, those rules are responsible for stopping about 2.5 million requests per day, none of which were from a human (I'd know, human visitors have very different access patterns, even when they visit the maze).
  If the bots were behaving correctly, and respected my robots.txt, I wouldn't need to fight them. But when they're DDoSing my sites from literally thousands of IPs, generating millions of requests a day, I will go to extreme lengths to make them go away.
- Alternatively, you could make your response more useful, removing the UI to aid the AI. After all, the user should be allowed to choose how they navigate the web.
  
  I am doing exactly that. AI turns my work into garbage, so I serve them garbage in the first place, so they have less work to do. I am helping AI!
  I'm also helping AI using visitors: they will either stop that practice, or stop visiting my stuff. In either case, we're both better off.

30 comments