Hi

Last week I published a blog post that I want to share with you, as I'm super eager to hear your thoughts on it as well. Do you recognize yourself in this? Here goes…

Ask AI to write something in a language I'm familiar with, and… well, I'm never super impressed. Yes, it works; yes, it needs fine-tuning and polishing. That's fine, though, it can be a huge productivity boost to help get you started.

Ask AI to write something in a language I don't know, and… my mind is blown 🤯. "How can it possibly be that something I don't understand, AI does so flawlessly??"

Now, taking that first experience into account, how big do you think the chances are that AI just happens to be a master in a technology I don't know, while it was average in a technology I do know?

Or… could it be that the quality of both contexts is the same, but I simply am more impressed by one of them because my understanding of that second context is much more limited? It's like a magic trick that's impressive as long as you don't know how it works.

What I find crucial to remember in this era of AI is that LLMs are essentially big averaging machines. They predict the next word based on what's most likely within a context. That is it. Don't get me wrong: we can do amazing things with simple technology; we can even turn something as simple as 1 and 0 into whole computer systems. The point isn't that AI is either good or bad, the point is that — by design — it will lead to an average result at best.

Of course, "average" — by definition — means that some people will be below average, just like me when it comes to using a language I don't know. That doesn't make the result objectively better, though. It's still average, overall.

Funnily enough, what I'm describing here has an official name: the Gell-Mann amnesia effect:

the phenomenon of experts reading articles within their fields of expertise and finding them to be error-ridden and full of misunderstanding, but seemingly forgetting those experiences when reading articles in the same publications written on topics outside of their fields of expertise, which they believe to be credible.

Let me know your thoughts!

From the feed

• TypePHP, Swoole's compiler for PHP, is now open-source
• Internals discussion to end PEAR endorsement
• "A" for "Apathy", another post in my "AI" series
• "A" for "Artificial", one more post in the same series
• Tempest 3.19.0 just released, with trusted proxies and database support for command buses
• I published the second episode of my "Return to RuneScape" series

That's it for this week, have a great weekend!

Brent