Thought Bytes

Watched this TED talk by Ken Robinson from 2006, and definitely worth revisiting as we’re now “redefining” intelligence in the AI era. Got thinking about connections between human intelligence and machines (LLMs) intelligence – different approaches but fascinating parallels.

1. “intelligence is diverse“

Ken mentioned that we think visually, kinesthetically, in sound, in abstract terms and in movements. This connects so much with what’s happening with Multimodal LLMs right now. Computer vision, OpenAI’s CLIP connecting text and images, robotics trying to learn kinesthetically… but here’s what I’m wondering – do humans actually need to convert between these formats? Or do they just naturally co-exist in our brains without translation? That’s something machines aren’t doing yet.

2. “intelligence is dynamic“

“Intelligence is wonderfully interactive” as Ken puts it. Yet our education system seems fixated on preparing people for specific jobs rather than nurturing this dynamic quality. Ironically, with AI, we’re actually pushing toward models that can do everything (the dream of AGI) – trying to build systems that can think fluidly across domains like humans naturally do. So why are we still educating humans for narrow specializations when even our machines are evolving toward versatility? Maybe we should be doubling down on what makes human thinking wonderfully messy and interconnected.

3. “intelligence is distinct“

Ken emphasizes how differently each person discovers their talents and carves their own path. Some find their element early, others take winding roads to discover what makes them unique. This individual distinctness is something our one-size-fits-all systems often miss. As AI gets better at mimicking general patterns, maybe the most human thing is how uniquely each of us expresses our intelligence and finds our calling.

Favorite Lines

One of my favorite lines in Ken’s talk: “I believe our only hope for the future is to adopt a new conception of human ecology, one in which we start to reconstitute our conception of the richness of human capacity.” (from 2006, but really needed right now!)

Questions for Another Day

IMO, we need to figure out what’s uniquely human that’s super hard for machines to replicate. But also – what can we learn from how AIs behave? Might be some surprising insights there too!