taybin

Great questions. The report itself says it doesn't know if it's intentional. Just that it does happen.

I mean, why would people collaborate or contribute?

The model doesn't have to perform unusual actions. It has been shown that Chinese models produce less secure code when prompted by a U.S. government persona. https://www.boozallen.com/expertise/cybersecurity/whats-in-a...

Decoherance. What happens to open source if everyone’s stock is different?

Undoubtedly lower. I’m okay with that.

"Claude write me a memorial for my employee. Make no mistakes. His life started when he joined my company."

People do weird things when grieving. This probably seemed like a wonderful tribute. From the outside, it seems tacky. But lots of things are tacky and memorials are for the living.

Eric's post was well written.

Also see: Slackware.

I always preferred Discordianism over CofSG, but both were cool.

Naw man, I'm taking this to my next gig!

You are right, but it is fun.

This was much better than I expected. Nice job.

Given that everything important has trade-offs, no, AI, even if it were very very smart wouldn’t be able to “fix everything”

Didn’t JWT have a similar thing, where you could specify the algorithm to use and that included “null”?

It’s not like it’s a lot of work to fill out permission slips. Probably easier than scanning the document. Receiving permission slips is the only way I’m informed about school trips anyways. I’m not sure this is a valid use case. Maybe for people who want even less participation with their families.

I have not heard of the Scots Wikipedia fiasco.

How’s that going to work when everyone turns one on and they’re all driving the profit margin to almost zero?

It’d be short sighted to not have a wrapper around which AI to use, even without implementation for them. Like, having an abstract parent class with a Siri implementation of it would be baseline good software engineering.

I think finding better systems with fewer false negatives and fewer false positives is very difficult and might even be impossible for many things. It might be my lack of imagination, but I often think about the measures taken to prevent child abuse. As I see it, you’re either going to get kids left in abusive situations or you’re going to have kids unjustly removed. Both are terrible. I’m not going to argue where the action bias should be. We live in an imperfect world.

I think preventing other crimes is similar. There will always be injustice on some side of the issue. And I’m not trying to argue that that means the status quo is great! But I don’t think we will eventually hone in a perfect system. Maybe just circle around some imaginary point, refining rules and processes until the whole thing breaks down.

Sounds great, but I'm not sure this will work well for individual contributors. Say the company is small, with a small team. Can you afford the amount of time it will take to find this bespoke unicorn engineer? Say the company is large, with a large team. Is this process going to scale to bring in, say, 50 new engineers?

Maybe it works better with management tiers where their impact is more indirect, and there are fewer of them, which mitigates the problems.

If I have a team in pain and we need more people on it, or different people, I don't know if I can wait six months for the one person I want to hire to become available and willing. Sometimes, you gotta go with who is available.

It does sound like you should have been investigated though. Like, this sounds like the (or a) system working. What would the alternatives be? Don't investigate for money laundering? Or have so much info that they can eliminate you as interesting without even having a visit?

Agreed, but it’s helpful for communicating to other musicians.

Now piano players will be able to join guitar players as not needing to read music or know theory!

I’m surprised that reading The Game hasn’t been added to the annual trainings for soldiers and DoD civilians.

Gaze upon my blogs, and despair.

I remember those. Trackbacks, right? Back when we were using MovableType?

Good. Kids need to learn how to think before letting other stuff do it for them.

This is exactly the same problem facing people trying to develop new music notations. In order to grasp the domain enough, they have to be experts in the existing music notation, and once you're an expert in it, the motivation to create something new goes away. From what I've seen, the people who want a new music notation are mostly people uncomfortable with sight reading.

Calling it a lie is pretty heavy.

You should see his book "Fluid Concepts and Creative Analogies, which is entirely about his experiments with AI in The Fluid Analogies Research Group with his grad students. Although I wonder if those techniques could be paired with LLMs somehow.

When I was a pretentious high schooler with fantasies of being an intellectual, I considered going there. I ended up not even applying. It just seemed too far out of the norm.

How is it dated?