I am an incoming physics undergraduate student at North Carolina State University.
joshua_mcmenemy
Karma: 0
Quality of work is definitely more relevant than ever at the moment. Duke University near me had a similar issue where one of the most published Scientists (Anil Potti), had epigenetic cancer diagnostic work go all the way to the clinic before other researchers figured out a lot of data had been removed to make the models work. Science has good write-up, on the study. It seemed like a similar case of pressure and wanting to confirm promising results led to a snowball of lies, much like Elizabeth Holmes at Theranos as well. I think in good cause work and competitive environments there is even more temptation to move forward before quality has been double checked.
I would push back a bit on your specific example of double checking code written by Claude by hand. Do you also expect people to read through the Python interpreter code line by line before writing Python scripts? I’m not pushing back against double checking, just that the way things are double checked needs to be scaleable. I don’t know the answer, but maybe it’s something along the lines of handwriting specific tests for the code you are generating.
I’ve definitely found myself talking ideas with AI more often, but as you mentioned it is definitely worth balancing with human conversation. I don’t think it is just “fake” acceptance of AI, but also the quick feedback and ability to immediately get information across fields, which is hard for one person to do during a conversation.
The Cohen Lab at Princeton has some cool papers on interacting with tissue’s bioelectric patterns to organize organoid tissue, which seems to intersect at a lot of the interests you described. The Lab website also has good visuals and high level descriptions if you don’t want to jump straight into a paper.
I also joined a community bio lab and went back to school. Working through physics problem sets by hand and carrying out experiments has helped balance out my perspective with more conceptual AI conversations. Plus, you run into people who may push back your ideas more than AI. Or, at least force you to explain more concretely and from different angels.
My perspective has been you have to get into spaces before they are mainstream to find more authentic signals. Web3/crypto was a super engaging crowd when it was a small group of nerds working to build an alternative to the banking system. The only people who were using terminal to download opensource code and set up wallets for tokens that were worth pennies, were people doing it for interest or strong beliefs. The same was true for gene editing before CRISPR or protein folding before AlphaFold. Basically, metrics and rankings tend to pop up more in crowded and well known spaces to deal with organizing everyone trying to get into the space.
Some metric mainstream optimization (salary, GPA etc) is worth it just to have a base foundation to navigate into spaces, but over focusing on it is definitely counter productive to finding interesting subcultures before they devolve or become overcrowded.
Great post.
The gaslighting of Alice can go even further once conflict emerges between Alice and the embedded system. Aaron Hillel Swartz is an Alice archetype that cared deeply about open software and published research being accessible. He spent years on committees at top universities like Harvard to make progress. When he didn’t get enough traction he ended up downloading a bunch of paywalled papers and making it free on the internet. This led to arrest and journal publishers winning a court case where he faced 35 years in prison and $1-million in fines. He eventually committed suicide.
Your post covered the passive and lazy resistance to change in-depth and did mention the collusion and malice to maintain status quo. However, I think the paranoia that emerges from Alice types is not entirely irrational given how much current parties may rely on a norm not changing to maintain current influence.