dwg465

Right now, we've got the Chat interface and the Copilot interface (a la github copilot) for Large Language Models.

What comes next?

Anyone know of any SF AI meetups? Seems like AGI/Alignment is a hot topic and would be interesting to get some discussions going.

I imagine we've got a city full of people interested/scared/motivated to think deeply about this and that maybe/possibly (if there's any optimism to be spared) could do something about it.

Does anything like this exist already? If not, anyone down?

Similar to the Maker Bill of Rights, what are the principles for building reusable software that would form a Developer's Bill of Rights?

https://cdn.makezine.com/make/MAKERS_RIGHTS.pdf

Hey HN,

My team recently started investing in optimizing Web Vitals (https://web.dev/vitals/) given Google's announcment that they will start impacting SEO rank.

We are noticing that our P75 metrics (LCP in particular) are very noisy, however at the P25 and P50 percentiles, these metrics are more consistent.

Has anyone noticed something similar or is it just us? If you or anyone you know at your org that works on web performance, I'd love to hear their thoughts :)

I worked for a larger tech company on the login / authentication platform and saw hackers cycle through 10,000+ IPs. How does a hacker acquire access to tens of thousands of IP addresses? Is it possible to fake an IP address?

From a light amount of research, Rigetti and IBM both developed their own instruction languages (Quil and OpenQASM respectively). Does each quantum framework develop an instruction language or are they standalone? It doesn't look like Q# or Cirq use a lower level language

I have some background in quantum mechanics and programming and am looking to get into quantum programming. There seem to be a lot of frameworks out there (QisKit from IBM, Forest from Rigetti, Cirq from Google, and Q# from Microsoft).

What are the tradeoffs with each? Does it matter which I learn?

After a quick search of Hacker News and a Google Search, it seems that the use Software Transactional Memory (STM) in Python was popular in the early 2010s but died out in the past 3 years. What is the current state of STMs in Python? What issues did researchers face?