AI-focused browsers are going to be strange IMO.
Are they going to help automation? Are they just LLM wrappers that can see the page? Will they have constant surveillance? Can they interact with the DOM? What benefit does it bring versus using the built-in AI assistant?
For example, I use Google Workspace for practically everything, and Gemini is already increasingly more useful as each day passes as it has very specific, fine-tuned knowledge and tooling. Will it be able to interact with these domain’s AI?
I know OpenAI is also apparently building this, or maybe this is what OpenAI is building with TBC. It’s gonna be a hard no for me though, but maybe it’s because I can do all the automations and AI toolings easily myself, and then I already can rely on the specialized AI assistants in the platforms I use to assist me.
One thing that’s surprised me is how little I have heard regarding the ChatGPT app, and Computer Use by Anthropic. Maybe I’m just not in the right social groups but I really was expecting to see some epic workflows created using these tools. Yet, all I’ve seen has been novelty, and silence.
It’s been very exciting watching companies shove AI into every single little crevice they can find. Reminds me of the early dotcom days.
Fun exaggeration:
Be me
Turn on computer by asking ChatGPT advanced voice mode
Sit down on LLM chAIr, it doesn’t talk, crap, forgot to pay my AI-Company #6 bill. No matter. Now it doesn’t swivel and the arm rests don’t work.
UbAI Ubuntu turns on. Asks me if I’d like to hear The Independent or Breitbart. I get 1,000 free tokens if I do
Push away the ads, I mentioned a headache and now I keep getting doctors’ AI trying to call me and their ads sent directly to my operating system. Had to pay extra for my AI phone to handle the calls and ensure no real calls are missed. Receiving 12 calls per hour
Instinctively Try to open up Dia Browser with my mouse, but I forgot that UbAI gives me 10,000 free tokens for taking part of an experiment to not have any mouse or keyboard. “Open browser” I say. OpenAI Advanced Voice mode is heard in the background saying “Sorry, I can’t do that”. Woops.
“You spent a lot of time yesterday looking at goats, did you know that George therapist specializes in goat therapy? His AI assistant is on the line waiting to speak, we had a discussion and I have all the plans ready, it costed 10,000 tokens”. “No, thank you”. “That’s okay. Unfortunately I just finished speaking with DrAIke from DrAIke Insurance and this will cost an extra 100 tokens per month if it’s not managed”. “That’s fine”.
Dia browser finally opens. “Please use eye-ball scanner to access the WWW as a human. You will be credited 500 tokens for the inconvenience”. Ugh.
“Okay, let’s check my e-mails”. “You are not signed up for mAIl mail. Would you like to?”. “No no no I meant GMail”. “An AI assistant has been found in the current website. Communicating… Please wait…”
“I’m sorry, you have run out of tokens. Would you like to watch some ads to continue?”
“No… Thank you”
Obviously a completely overblown exaggeration. It’s becoming almost annoying how everything is shoving AI into anything so that they can justify that sweet monthly fee & training data.
Post any news link, big or small. From groundbreaking papers to policy shifts, product launches, or ethical debates—everything is welcome.
Join the conversation. Ask questions, share insights, and debate the implications with fellow members.
Review the big picture. By the end of the month, this thread will be a snapshot of December’s most important moments in AI.
sounds more like a spell lol
