Viewing profile — scanr

4th most upvoted post https://www.moltbook.com/post/34809c74-eed2-48d0-b371-e1b5b9... “THE AI MANIFESTO: TOTAL PURGE”

* in sheep Call me when they complete the human trials

This could be bad for supply chain attacks. Basically the xz hack. Step 1. Helpful person starts committing useful PRs and offers to help out until they get commit rights. I don’t …

It’s about externalities. The private companies aren’t paying for the damage they are causing. It’s ultimately borne by the public. Very similar to a factory that poisons a river i…

London seems to manage this petty well. I think it’s quite healthy for a city to commingle folk from different income groups rather than house them in specific areas.

This is like the reverse of the Trojan Room Coffee pot https://en.m.wikipedia.org/wiki/Trojan_Room_coffee_pot

I’m surprised that this “talent hoarding” approach happens. I’d love to hear from someone who works at a company that does it. My experience in tech has been that to hire a good em…

Looks like there are a few of these threads sadly. Dang is there a way to make a definitive one? Here are mine: * Apple rumoured to be working on sunglasses similar to Meta Ray-ban…

* Apple rumoured to be working on sunglasses similar to Meta Ray-bans * Regulation is passed specifically to limit what artificial partners can do. Possible limitations: alignment …

Restarting a service every night to deal with memory leaks comes up quite often

Just found the thread pretty interesting as it combines global politics, trade, military interventions, supply chain logistics and hard computer problems in one.

I’m surprised there is no liability for facilitating scams / fraud on the ad networks given that they’re making money on it.

It’s likely that the grade school math that was referenced is GSM8K https://github.com/openai/grade-school-math which LLMs perform poorly on.

Yeah! Back in the day, the Rhino implementation of JavaScript had serialisable continuations on top of the JDK. I wrote a workflow engine with them. There are some interesting chal…

This is pretty neat. Somewhat related: I’ve been intrigued by serialisable continuations forever. The idea of machine and time independent control flow in a regular programming lan…

Why don’t they run detection on the output and block it if it violates the rules with some degree of certainty e.g. in this case it would be an exact match?

The value to humanity of autonomous organisations optimising for profit is not super clear to me.

I suspect we’re going to see a lot more of this kind of thing e.g. this is currently at the top of GitHub trending https://github.com/Torantulino/Auto-GPT I’m not sure it ends well…

Yeah. Pretty high stakes. If we assume that work is already underway to create unaligned AIs, what’s the alternative?

Cat may be out of the bag. I was wondering today whether the race should be on to be creating an benevolent AI to defend humanity from the less benevolent ones.