Talkative Tuesday 🤓 ➡️ ~ Reflective Intelligence

 

I'm a really big fan of Artificial Intelligence (since long before it became trendy), and I’ve always been fascinated and impressed by its recent advances and developments. However, there’s one part of AI that I personally feel hasn’t been fully explored. I’m not entirely sure of the proper term for it, but I like to call it reflective intelligence.

Essentially, the idea is that an AI model could study a human subject—how they make decisions—and search for logical patterns in those decisions. The AI would then use those patterns to make choices of its own, ideally replicating the decisions that specific person would’ve made under the same circumstances.

In the TV series The Flash, Cisco actually pulled something like this off. He created an algorithm called The B.A.R.I. (Barry Allen Replicated Intelligence—brilliant name, by the way). It studied Barry’s decision-making using all the data Cisco could gather, and it used those patterns to make new decisions. The algorithm was so good that when it was tested, it correctly answered a question about Barry that even Barry himself got wrong. The question? His favorite movie. Barry answered The Empire Strikes Back, but B.A.R.I. answered Jurassic Park—and Barry had to admit that yeah, he actually did prefer Jurassic Park. 🤯

Of course, this is a fictional (and exaggerated) example—there ain't no damn algorithm in the multiverse that knows my movie preferences better than moi 😤—but I do think something like this is possible to a certain extent.

Take chess bots on Chess.com, for instance. Even when bots have similar or identical ratings, they can differ significantly in playstyle. That’s because they often mimic distinct decision-making patterns. Chess.com even creates bots based on real players and influencers to replicate their playstyles. But interestingly, when those same players face their AI counterparts, they’re usually disappointed—because even though the bots have matching ratings, they don’t play like them.

And that always made me wonder—why not?

Chess.com already has a record of all your online games linked to your account. Couldn’t they just feed your recent games into a logic analysis engine and—voilà!—a bot that plays exactly like you?

For me, chess is the perfect place to begin exploring reflective intelligence, and here’s why:

👉    Chess is a two-sided game, making it easier to observe the situations (or positions) and how a player responds to them.

👉    It’s one of the most complex two-player games—unlike tic-tac-toe or mancala, which are “solved” games, chess remains impossible to master to a mathematically unbeatable level.

👉    Because of this, players develop unique decision-making styles and patterns over time—patterns that can be studied, mapped, and potentially mimicked.

So why hasn’t this already been done?

If we explore and develop reflective intelligence thoroughly, I believe it would bring some neat advantages. On a small scale, it could help make decisions on someone’s behalf in their absence—mirroring their thought process as closely as possible. On a larger scale, it could open new frontiers in analyzing human philosophical, conscious, and subconscious psychology in ways that would otherwise be impossible.

So yeah—thanks for joining me for Talkative Tuesday 🤓. I hope my little rabbit hole of an idea left you curious. Have a lovely day! 👋