Mastodawn

Tip: if someone in a work context tells you "here are some suggestions I got from a chatbot", just replace "chatbot" in your head with "my preschool kid".
Most likely the suggestions are just as useful, but somehow your colleagues would never dare to suggest you should follow advice from their preschoolers.
#workwork

Show thread

Andy Randall 🇺🇦Apr 23

@mxk great sound bite in a social post and this used to be true — but not with the latest generation of chatbots (paid subscriptions, not the free tiers). I regularly use copilot for suggestions - eg “review this email/doc and give me feedback on how it will be received by my target audience”; or “how would our compeititors respond to this?”. I get nuanced actionable feedback at a quite advanced level, better than many humans I work with tbh.

Show thread

Andy Randall 🇺🇦Apr 23

@mxk you’re entitled (and probably justified) to hate AI for many reasons, but to ignore its very real capabilities is intellectually dishonest

Show thread

mxk Apr 23

@ahrkrak AI might be a useful tool, but telling someone else, who probably is knowledgeable in their field, just what an LLM told you, still is offensive and probably a waste of your colleagues time.

If you are using LLMs, it just should be your part to validate their output to a point where you can represent the suggestion in a way that you can own and defend it.

@mxk @ahrkrak 💯

@mxk not so sure, today I much rather had dealt with pre-schoolers than with the mob that showed up in my meetings, and they did that just fine without AI 🤪

Show thread

Clemens Apr 23

@mxk Yeah, there should be pushback against doing that, it scrams "my time is more valuable than yours".

I don't know how to make that happen, though. In the past I've actually spent the time to review some of those suggestions and provided detailed feedback about why the suggestions are all wrong, and that seems to have caused enough shame that the person hasn't done it again since, but this doesn't scale.