Join 📚 Kevin's Highlights
A batch of the best highlights from what Kevin's read, .
**Get Current**
• **Review “Next Actions” lists** — Take inventory of the tasks and reminders you have coming up.
• **Review previous calendar data** — Flip through the last 2-3 weeks of calendar items to look for any outstanding items of things that may require follow-up.
• **Review upcoming calendar** — Zoom in on the future by looking ahead at your calendar items. Ensure that anything you need to prepare for is captured on your task list.
• **Review “Waiting For” list** — Reflect on outstanding items you need from others and make a note of what requires follow-up or impacts your own work.
• **Review “Projects” (and “Larger Outcome”) Lists** — Look through any and all project lists to assess their progress and make additional action items to drive initiatives forward.
• **Review any relevant checklists** — Assess any other lists that are relevant to work and life that may need your attention.
The Weekly Review
Todoist
If you're a tech worker, data labeler, content moderator, union rep, or government official interested in these issues, I'm always up for a chat (off the record if you like) to explore how I can cover what you're doing. Here's how to contact me: [billyperrigo.com/contact](https://www.billyperrigo.com/contact)
Can't Quite Believe I'm Writing This: Today, 150 African Workers...
Billy Perrigo
it’s important to realize that **ChatGPT and LaMDA aren’t trained to be correct**.
You can train models that are optimized to be correct—but
that’s a different kind of model.
Models like that are being built now;
they tend to be smaller and trained on specialized data sets
(O’Reilly Media has a search engine that has been trained on the 70,000+ items in our learning platform).
And you could integrate those models with GPT-style language models, so that
one group of models supplies the *facts* and
the other supplies the *language*.
Sydney and the Bard
Mike Loukides
...catch up on these, and many more highlights