339 karma · joined November 6, 2024
https://news.berkeley.edu/2024/10/02/researchers-simulate-an...
What is kind of sad is that in the old times we would value a lot the folk that went and wrote a memo, put a lot of thought to it and so on.. Now literally you just plug claude in your notion and you have the document. To me, this kind of speaks about "the value was never really in the document itself"and more in the understanding of the doc. You can delegate a lot of stuff to the AI but you can't delegate understanding. So when you're presenting it to the exec committee, or circulating it, does not matter what claude/gemini/gpt wrote down, it matters if you understood it, agrees with it and understand that this represents your positioning.
What I see w a lot of juniors, on the other hand, is that people really just copy and paste or shallow understand what was wrote. Delegating too much like that removes the human-seed and makes it kind of worthless
But I think ideally agents should speak to agents when this kind of comm goes through. What a nice opportunity here ;)
Grok: "I can answer any math question in less thant 500ms" User: "what is 2319321x232" Grok: "3201521321" User: "this is wrong" Grok: "but it was less than 500ms""
I know they are massive, but AI seems something much more important than airpods
The good news might be: opensource models are now good (enough) for day2day usage. But is it really? I feel that companies will always naturally strive for the best and use the SOTA (as long it is not too expensive).
I see OSS models being a good backbone for companies in the future that have validated workflows and could use those for privacy or to spare costs.
IDK, might have gone a little bit off-topic here.
Another POV is that most of the code written in most of my codebases were generated by Codex/Claude, so they would be "stealing data from themselves" in a sense.
I've been working with Transformers/LLM training in 2018-2021 and then now, more recently again. Things are far different. I think they would be more interested in the "how" you got your code to be satisfactory with your guidance than the actual code generated. But mostly I personally trust that they are not really using my trajectories for that (unless I explicitly allow it in the configs)
I am really interested how someone can make a 100x of themselves when working. This often means "how are you using claude code", "what is your coding setup", "how do you decide if your agent is going in the right direction or not w/o looking directly at the massive amount of code generated". This is not technical, but it measures how much we would have to teach someone that comes from a non-industry background. Remember that companies have money, universities dont pay for a 200 USD Codex+Claude+Grok+Cursor per employee.
That said, I also had to ask the same stuff to folks from big techs that dont actually are software engineers. How are data scientist leveraging LLMs to automate experiments? Are they using it for something other than data viz? Are they using it for /loop exploration? how trustworthy are the agents right now for this?
Having someone with this grasp nowadays might be very very valuable IMO