stevenwliao
Joined 11 years ago
Joined 11 years ago
I'm fairly certain the error rate for all those shops is less than 5%.
I only have a limited amount of time to read. Skipping someone's Internet comment because it looks like spam often means I get to engage with something else.
Summary here: https://simonwillison.net/2025/Apr/11/camel/
TLDR: Have two LLMs, one privileged and quarantined. Generate Python code with the privileged one. Check code with a custom interpreter to enforce security requirements.
Screenshot: https://imgur.com/a/2tLoFkQ
IIRC the rates are not competitive compared to local plans though, so not sure if it's worth it for you.
Plenty of AI models cite their sources.
(Assumptions: I watch 10 hours of YouTube videos per week, YouTube shows 30 seconds of ads per 10 minutes of video)
Which is the more likely scenario?
1. OP buys bits and throws them at their favorite streamers despite not being willing to pay for subscriptions.
2. OP pays nothing for Twitch and doesn't watch a single ad.
Where is the precedent for this?
> He was chief technology officer of Facebook, and also served as its vice president of engineering, until 2008.
> D'Angelo was an advisor to and investor in Instagram before its acquisition by Facebook in 2012.