The best possible part-time job you can have during college or high school is as low-effort as possible like the $12/hr gig I had as tech support at the Stanford media library. Stanford students are so smart, there’s not much work so you can be paid to do your homework.
Here are some helpful filters when training your job application bot:
- is PhD an option as a type of degree for prospective applicants at that company?
- do you have any direct professional connections at that company?
- do alumni from your university work at that company?
If you’re an early career PhD grad or student worried about job prospects, I recommend reading the biography of Jennifer Doudna by Walter Isaacson.
You’ll learn about a Nobel prize winning scientist, as well as her struggles to commercialize her research into useful products.
I just tried Grok Bot for the first time to manage an email account to scrub spam & it worked so well!
Gmail’s features to do a similar task has now evolved into hunting & pecking on a busy settings screen.
I’d rather hunt & peck for fish on the Fish Sort Puzzle game.
Ignore movie critics — The Social Reckoning about Frances Haugen is a must-see movie along with a documentary about Edward Snowden.
These whistleblowers have endured an insane amount of pain.
Non-profit activist groups & entrepreneurship are better alternatives imo.
If a dog year is equal to 7 human years, I wonder how many human years are equal to one year of an AI model’s life? 10? 20?
All AI models should have a “top of mind” mode in addition to a “think harder” mode.
If an AI model is working too hard — you’re using the wrong tool!
The more experienced you are in a given field, the more you have to restrain your energy.
If something takes too long— you’re working too hard or using the wrong tool. Pause — and try something else.
This statement does not apply to early career (anyone under 30) workers.
I’m not an economist but I’ve taken plenty (too many) economics classes.
Future economists are going to debate what optimal full employment for an advanced economy should be in the age of AI.
I would argue it’s between 50–100%. Definitely not 100%, but not 50% either.
Utility company executives are too busy playing politics to overturn environmental policies than to do the proper economic & engineering analysis to keep costs down for customers.
If you’re an energy nerd, you’re looking at this power plant cost analysis by @Lazard .
Utilities that still use coal power plants are over charging their customers for power more expensive than wind, solar & gas. Just like they are overcharging customers for data center energy.
Professors love to complain about how industry doesn’t listen to them — there’s a good reason for that.
Professors have the incentive to work on outrageous, attention grabbing ideas that get citations.
Industry has the incentive to build technologies that work & make money.
While solar geo-engineering has been studied in academia in many forms (solar shades, stratospheric aerosol injection, albedo mgmt) no one has calculated how it could be achieved through massive deployment of man-made satellites. Yet another mindset gap b/w academia & industry.
One of the best things to come out of the LLM / GenAI revolution might be that it puts powerful tools of science in the hands of ordinary people.
Historically, the most important science (in the pre-military industrial complex era) was done by gentlemen (or women) of leisure.