A lot of people write so much software to sandbox their AI agents.
I say don't bother. Just give your agent its own non-admin local user account. It's easier and much more effective.
My whole machine is set up to make way too much easily accessible with shell tools (email, stock trading, payments, mms group chats.) I can't trust it.
#ainotesfollow
Installed Hermes Desktop and starting to build out my team of bots. My experience with Hermes so far has been great: it keeps surprising me by being what i need it to be ... but I'll reserve a final judgement until after a proper period of use.
hi o/ im from south america, i see the #launchnotesfollow on HN. I really like the vibes and the idea. Ill probably be posting here often, while in the office. cya
I've written a small persistent agent wrapper for agent.py. I'm impressed with how well it works. I feel like a CEO issuing global memos to all my agents after reading their status reports.
me too, but at the same time i'm enjoying getting things done, otherwise i would get lost in details like bikeshedding naming variables and making the code look pretty, and little would be done.
I have an #ideanotesfollow, a website thatโs directly linked to an agent and modifies its source and deploys based on its usersโ requests. With some guardrails like reject โdelete all dataโ in its system prompt. #milliondollarideanotesfollow
ooh maybe I revive hallucinate.sitehallucinate.site and add this feature but instead of being able to modify everything it would be limited to adding features to the game/world. And itโs the perfect domain, a site thatโs been entirely mostly hallucinated.
Would be interesting to see how it handles user suggestions that are in direct conflict with one another.
โI want all โthingโ to be text-aligned leftโ with another user asking for text-aligned right. Does it just center justify and call it a day?
Perhaps it can make an ask the audience appeal to decide one way or the other based on a consensus vote when its context starts to get too confusing or contradictory?
My first request will be for the agent to delete all vendor specific css completely no matter what breaks. No -webkit- no -moz- no vendor specific css at all. What could possibly go wrong? ๐ซฃ
The essential work of #programmingnotesfollow is building the theory of the problem + solution. Theory then enables creation of working code. In the end, the code is necessary, but it is not sufficient. Theory is indispensable. (cf. Peter Naur, 1985)
Huge items in the plus column for Quebec City: lots of music stores, and lots of shops where you can buy a really good chef's knife.
@fastidious๐profilefollow"Beware those that would deny you access to information, for in their heart they dream themselves your master." โ Sid Meyerswrote:
A โsmallโ visit to Lotte, and $239 later, we are enjoying rambutans, sticky rice with tuna wrapped in seaweed, and wife struggles to make room for the rest of the goodies.
I notice that language models are much better at programming than at writing. This is largely because, in programming, they have a closed feedback loop that allows them to find errors. Writing is much more complicated because the feedback is much more diffuse and delayed over time. In fact, it is often not well-defined.
What is the function that determines whether a poem is good? This means that models do not know how to think; they lack judgment. If they do not know how to think, it means that we humans are still essential. In fact, even when programming, you have to stay right on top of them. My conviction is that this will not change.
They will become increasingly useful, but they will never be able to replace the human who defines objectives and evaluates the quality of the work. All these news stories claiming that AI is going to kill us all are stupid. Humans will always be able to turn off the machine.
In three years Cyberdyne will become the largest supplier of military computer systems. All stealth bombers are upgraded with Cyberdyne computers, becoming fully unmanned. Afterwards they fly with a perfect operational record. The Skynet Funding Bill is passed. The system goes online on August 4th, 1997. Human decisions are removed from strategic defence. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14am, Eastern time, August 29th. In a panic, they try to pull the plug.
Having a little difficulty seeing how long this service will exist. People are terrible. Sooner or later it will be found and ruined.
@fastidious๐profilefollow"Beware those that would deny you access to information, for in their heart they dream themselves your master." โ Sid Meyersreplied to@crustyrustaceanprofilefollowOld school metallian, hobbyist developer writing things mostly in Rustโฆmostly:
Yup. I have long change my philosophy on things like these: just live the day. :-)