Jul 12, 2023Liked by Marija Gavrilov, Chantal Smith
Given the first chart, what's the implication of Jevon's Paradox on software development? Less skilled people can do higher value work, so they shift to these jobs. What next? What's the implications to other structured creative sectors - like planning, construction, marketing...
It remains to be seen, but the paper we link to about implications on data scientists offers ideas for what the shift could look like for that profession.
Jul 12, 2023·edited Jul 12, 2023Liked by Marija Gavrilov
I was listening to an "emergency podcast" yesterday with programmers geeking out about Code Interpreter. Detailing experiments, and use cases in a colaberative manner, they appear to be using it to automate the parts of programming they don't like and to refactor code, etc. People who have been programming for 25yrs were learning new stuff because of this, the zeal and excitement was palpable. This is happening now. Programmers not using this in 6 weeks or so are going to be left behind.
I was also reading an interesting paper all about applying personailty theory to large scale LLM's to see what they were like as people on the Big Five personality scale. They were higher accross the board than the average human, except in openness to new experience, where they scored lower on average than humans, they were especially far more neurotic than humans which I find odd. Not read the conclusions yet.
Given the first chart, what's the implication of Jevon's Paradox on software development? Less skilled people can do higher value work, so they shift to these jobs. What next? What's the implications to other structured creative sectors - like planning, construction, marketing...
It remains to be seen, but the paper we link to about implications on data scientists offers ideas for what the shift could look like for that profession.
I was listening to an "emergency podcast" yesterday with programmers geeking out about Code Interpreter. Detailing experiments, and use cases in a colaberative manner, they appear to be using it to automate the parts of programming they don't like and to refactor code, etc. People who have been programming for 25yrs were learning new stuff because of this, the zeal and excitement was palpable. This is happening now. Programmers not using this in 6 weeks or so are going to be left behind.
I was also reading an interesting paper all about applying personailty theory to large scale LLM's to see what they were like as people on the Big Five personality scale. They were higher accross the board than the average human, except in openness to new experience, where they scored lower on average than humans, they were especially far more neurotic than humans which I find odd. Not read the conclusions yet.
Very interesting! Mind sharing the links to the paper and the podcast? Thanks
https://open.substack.com/pub/swyx/p/code-interpreter?utm_source=share&utm_medium=android&r=qoyq
https://arxiv.org/pdf/2307.00184.pdf
AI/LLMs today seems like a 21st Century version oh a Nantucket sleighride—except we are being towed by whales and unicorns.