Posted by yks 15 hours ago
I can't even begin to comprehend what kind of personal and PII data this thing will be scraping from its users. Caveat emptor!
How do you have that?
I dont know they trust me because they are stupid.
The fact that this conversation is public and anyone uses meta garbage is what's wrong with so many things
I don't get it. I think he meant exactly how it looks on paper
Grok Bot doesn't have direct access to Twitter like Grok on x.com does, for no good reason, but it does have an API plugin which gives you free $20 credit if you install it.
When these companies want to lock you into their platforms it seems kinda dumb to not give their own products access to their own products.
All of Meta's products have to be user-hostile, because the first thing you want is interop and that kills their monetization.
For this kind of tool to be useful it basically has to be able to impersonate you digitally. I know I personally have zero trust for Meta to do this securely and without exploiting me, whereas Google I feel a lot better about here. After all Google has made a business out of safeguarding data with gmail, drive, etc. Same with Microsoft and Apple. Meta, OpenAI, not so much.
On the other hand, I’m always surprised how little many seem to care about data privacy.
It's certainly honest.
Is this marketing to end-users or to stakeholders?
The Agent Computer is super powerful, but by definition its easier than ever to build one yourself, and a the result is a personal and private computer.
I know most "normal" people wont want to build their own this but there are still a lot of options that don't involve giving Meta or Grok all your data.
I put some notes on all this here: https://housecat.com/blog/agent-computer-101
It can do file organizing and wrangling when needed, it has simple web access with the curl tool, it can analyze media file contents. It's analyzed my local data and written several little tools (cli, TUI, GUI) to manage it automatically or give comprehensive reports. I can create new media through ComfyUI and Krita AI.
Sure, it can't book a flight or organize the family calendar for the next month, but I don't have any need for those.
edit: llama.cpp LLM model and ComfyUI image/video models can't live in the VRAM at the same time, but I let it code a little KDE widget with two buttons: one unloads the llama.cpp model and another one unloads the ComfyUI models.
why can't it do that? just run a VPS inside WSL or another container inside your PC. give the agent that sandbox.
In 2015 Facebook (not yet renamed to Meta) launched a personal assistant named simply "M" that lived in Messenger and functioned remarkably like ChatGPT does today. At the time it was basically entirely humans on the backend, with the idea that they could replace humans incrementally as they collected data. Really ahead of its time. They shut it down in 2018, which seems shortsighted in hindsight. That data would have been invaluable just a few years later.
https://www.technologyreview.com/2017/04/14/152563/facebooks...
2015 story: https://www.wired.com/2015/02/magic-the-startup/
Thread seeking a follow-up in 2022: https://news.ycombinator.com/item?id=34148402
If Muse could actually just show me content that I want to see and cut out all of the other crap, I'd be willing to use it. Of course, there's no way Zucc will pay for an expensive AI service that will undercut his main revenue stream.