In 2011, Nobel laureate Daniel Kahneman proposed two systems of thinking - “slow”, deliberate thinking. And instinctual, “fast” thinking - which paradoxically "costs" you more.
I think we need to apply a third category… Thinking with AI.
AI has enabled us to think even faster.
But often times, slow deliberate thinking cuts costs in comparison...
Let me explain.
Anyone who as applied themselves seriously to AI has gotten into that tricky situation I’d like to call a “prompt spiral” - when prompting yourself out of a problem wastes even more time than actually thinking through it.
We’re already seeing this with vibe-coding where developers speed through AI credits instead of actually thinking through the issue. Costing their companies even more.
This points to the main gap in AI investment - the infrastructure between your ears.
Therefore, the goal of universities should be inverted:
How can we enable students to think - so we save time / money using AI?
Not enough investment is being made in “prompt engineering” or prompt craft.
The tricky thing with AI is there is no obvious cliff to which investment in AI jumps off of (unless perhaps you’re thinking of China).
The loss in revenue in AI is the "time lost" - days, weeks perhaps when teams of individuals could have just gotten their noggins together and critically thought themselves out the problem instead of working in silos with their so called new "interns".
Prompt engineering - knowing how to prompt, or more crucially when NOT to prompt at all, will help business executives and tech workers apply themselves on the one thing wasting their time.
That chatbot thinking too fast.