@rayofhk: What if AGI doesn’t eliminate scarcity — but simply moves it somewhere else? Much of the AI race has been organized around one major bottleneck: intelligence. Build better models. Train them with more compute. Make them reason better. Make them more capable. But suppose we really are entering an era in which increasingly powerful intelligence becomes abundant. What happens next? More capable intelligence creates more agents. More agents create more inference. More inference requires more compute. More compute requires more chips, memory, data centers and electricity. More autonomous action requires more permissions, cybersecurity, reliability and trust. And greater economic importance creates more questions around capital, regulation, governance and responsibility. This is why I find the idea of AGI particularly interesting through my Six-Layer AI Ecosystem Model. Improving Layer One — intelligence and applications — does not make the layers underneath it disappear. It can increase the pressure on them. And that means solving one bottleneck may simply reveal the bottlenecks behind it. If intelligence becomes abundant, perhaps the next scarce resource is compute. Or electricity. Or memory. Or infrastructure. Or permission. Or trust. Or human judgment. Or perhaps the ability to coordinate all of them. Scarcity doesn’t disappear. It relocates. So perhaps AGI isn’t the end of the scarcity story. It may accelerate the migration of scarcity. And that leads to a question I think will become increasingly important for both technology and investing: If intelligence becomes abundant, where does the next bottleneck appear — and who controls it? Because the arrival of AGI may not eliminate bottlenecks. It may expose them faster. #AI #AGI #aiagents #semiconductor #raysixlayeraiecosystemmodel

🐉Ray🦅Sep20-200DayAnniversary

Open In TikTok:

Region: SG

Saturday 05 September 2026 10:14:09 GMT