The Weekly Prompt #2: The Good, The Bad, and The Hallucination

Hi friend,

Welcome back.

Last week I mentioned 'Steelmanning' your own ideas. A few of you tried it. One person said it completely changed a decision they were about to make. That's exactly the point.

This week, something worth knowing:

What's happening in AI right now:

A new Stanford report dropped this week with a stat that stopped me mid-scroll. Generative AI hit 53% population adoption in just three years, faster than the personal computer and faster than the internet. Think about that for a second. A technology that didn't exist in most people's daily lives three years ago is now used by more than half the world. Stanford HAI

And yet, only 6% of teachers say their school's AI policies are actually clear. Stanford HAI

Access is everywhere. Understanding is not. (Sound familiar?)

Also this week: a Harvard study found that an AI reasoning model outperformed experienced physicians at diagnosing ER patients using only electronic health records, correctly identifying conditions at a noticeably higher rate. I'm not saying AI is replacing doctors. But I am saying the gap between people who know how to use this well and people who don't is getting harder to ignore. Crescendo AI

Something to try this week:

Ask AI to explain something you're working on, then follow it up with this:

"What are the three most important things I haven't asked you about yet?"

It's a simple prompt. But it forces the model to think about what's missing from your question, not just answer what's in front of it. You'll be surprised how often it surfaces something genuinely useful that you didn't think to ask.

Quote of the week:

"The question is not whether machines can think. The question is whether people can." — B.F. Skinner

See you next week,

Shantanu

P.S. If someone forwarded this to you, you can subscribe here www.auracle.co.nz. Worth it (Allegedly).