Advice from a guy who will be paying taxes to keep the Mamdani Caliphate running: “I would ask: “May I meet you?” before engaging further in a conversation. I almost never got a No.”
Gemini applying this for Mindy the Crippler:
ChatGPT does a comparatively crummy job:
Grok is off in its own world (I could get it to use this puppyhood picture of Mindy the Crippler, but it simply ignored my request to use the one of the golden retriever standing on the windowsill of the minivan):
The mane reason our modern world of AI garbage looks so much worse than what we had 3 years ago is we’re now diluting snippets of natural language into a lot more information. Computer graphics previously never looked as good as practical effects because they diluted information.
Would be funnier if it looked like Mindy made the sign herself, like with a sharpie and a square of fabric from a postal workers pants.
I (M59) would probably respond to “May I meet you?” from a woman (as unlikely as that would be) with “I don’t know, madam…do you really want to take on that burden?” as a cute ice breaker. If it was anyone else, I’d run for it. On the other hand, if I asked someone, I probably would get pepper sprayed.
I saw a fun new AI app today being advertised on Reddit called Leetcode Wizard for cheating on FAANG fizz-buzz interviews using a hidden session. “AI, a thousand and one household uses!” None legitimate or productive.
As I was strolling the grounds today, it occurred to me that FAANG should openly allow AI during interview questions anyway, since it is shoving it down the rest of our throats as a productivity tool — eating your own dogfood and all that. Also, it is likely the manner in which an employee who cheated his way through college is going to do his/her/xher work anyway. As an interviewer, I myself would probably give the benefit of the doubt to someone who skipped AI and got a wrong answer, over someone submitting something from AI that looked right but had a hidden bug. If they insist on not allowing AI, maybe they do know something about the efficacy of AI and the wisdom of P.T. Barnum.
The purpose of the coding interviews is not to test the exact skills that will be used on the job, but rather how well someone knows the basics, whether they’re willing to grind long enough to pass an arbitrary test (screens for people who have attitude problems or are otherwise not psychologically suited to working for a large company), and if they’re smart enough to do the job. Testing how well people can use AI doesn’t do any of that. In an age of AI where companies theoretically need fewer employees, there will coincidentally probably also be fewer people who are able to pass a coding interview without using AI.
@Ryan
Thanks, think I get it now, they are looking for general intelligence and cultural fit for doing pointless busy work and brown-nosing. Clearly that is not unambiguously determinable with a $500,000 degree from an Ivy. After the AI Winter Pt. Deux, they are going to need 100X engineers to clean up the mess in an AI-free, air-gapped Faraday cage. But still…maybe they can add on an wicked-mad AI skillz test to separate the leet wheat from the less-leet chaff for people that are so smart they can get AI to do something useful, given the lower staffing levels.
P.S.
> screens for people who have attitude problems or are otherwise not psychologically suited to working for a large company), and if they’re smart enough to do the job
Attitude problems, ill-suited psychologically, and dumb were all dings on my final performance review in the Fortune 500. I guess screening interviews aren’t always 100% effective — those doesn’t readily show, in my case, does it?