Hacker News Reader

Comment by treis | Hacker News Reader

This is like people railing against images in their webpages or that run their browsers without JavaScript. Like that LLMs are a fundamental paradigm shift in how we interact with computers and the web. How it all plays out is up for debate but there's no doubt that 10 years from now LLMs will play a role in nearly every interaction we have with a computer. The only way that's not true is if something better comes along.

Pining for pre-AI world is like wanting families to gather around the radio. Those days are gone.

View on HN

> there's no doubt that 10 years from now LLMs will play a role in nearly every interaction we have with a computer

I tend to agree with you. Doesn’t make what Mozilla is doing sensible.

In 1995, one could correctly observe that the internet would “play a role in nearly every interaction we have with a computer.” It would not follow that every app must reïnvent the network stack.

An AI helping out can be useful. Every app being a tiny AI is a cacophony of idiots.

If they were training their own foundational model your analogy would be on the mark. But this is more like porting a command line application to a GUI.

> this is more like porting a command line application to a GUI

CLIs preceded GUIs. This would be like a CLI jamming mini GUIs into its flow because that’s the next thing.

Isn’t that called a TUI?

That’s a good analogy. My response would be how many of those products stuck around as TUIs versus CLIs or GUIs?

No one's "Pining for pre-AI world". Many people use some amount of AI every day, whether they know it or not. I use Claude Code extensively, for example.

I know it's a paradigm shift. That's not the problem. The problem is that it's often wedged into workflows in ways that aren't helpful to me or are actively harmful. And then there's the question of what is done with the data. I don't need another tech company, non-profit or not, getting a hold of my chatbot conversation history and doing God knows with it.

Mozilla should be a better facilitator of the ecosystem around AI than just putting it in Firefox. Take care of the concerns before just shouting "me too" on a bunch of LLM features, which, to be honest, shouldn't even be a concern for FLOSS.

If someone wants GenAI in Firefox, they can create a branch, design it, implement it, and put it up for discussion. I don't need some CEO telling me the direction of the project. It's the cathedral vs the bazaar, which has been a major part of the FLOSS ethos for decades now.

Five years ago, we could've read this same comment but with "LLMs" replaced with "blockchain" or "crypto".

Yes, it might totally be the case that in 5 years this comment reads as correctly predicting the future that is to come. But it's also possible that it doesn't.

It's not at all clear to me which things will persist in time at the moment they are getting popular. There are lots of technologies that look promising in the beginning and up fizzling out.

Browsers are useful now, and they have been useful for a while. It seems to me like a safer bet to invest on them still doing what they are useful at, in the case that the web keeps being a thing for a while still :)

No, it is nothing like people railing against images or JavaScript in webpages. Those are features of websites that the browser needs to support to provide the full intended experience to the user of websites that use them. In what way is integrating an LLM, let alone an agentic AI, needed to provide the full intended experience of which websites?

Why would my browser that is used to display static HTML need high performance 3D rendering or a low level assembly target?

The websites that use LLMs don't exist yet. Think of something like meetup.com. Instead of that you ask your LLM for something to do this weekend. It finds events, other people looking for something to do, and you have an ad hoc meetup.

Beyond that it's like looking at a blackberry and predicting what it'll be used for in the future. Hard to say "oh gay men will use that for casual hookups and it'll be called Grindr".

I think a lot of families actually restrict or eliminate screen time altogether, and gather round using either no technology, old technology, or new technology that imitates old technology, e.g. Yoto. There is value found in reducing our technology use. Some people may want to avoid AI on ethical grounds, or may be cautious about the cognitive effects of its use.

> there's no doubt that 10 years from now LLMs will play a role in nearly every interaction we have with a computer

There is a great deal of doubt about that. I think that's a very unlikely prediction you're making.

I can doubt that!!!!

Reply on news.ycombinator.com