Generic AI strategies say nothing — real decisions are what to change, unlock and stop
alex_verem · x · 2026-10-04
A widely shared enterprise AI playbook — fix data, identify use cases, measure ROI, add governance — is sound but indistinguishable from every other company's plan, argues the author.
Before engineering starts building, he says, someone must decide:
- Which operation is actually being changed?
- What becomes possible that wasn't before?
- What will the company stop doing once it works?
Example: an agent drafting billing follow-up emails saves time; a system that handles routine follow-up, records outcomes and routes exceptions changes how the team works. But the resulting capacity (more accounts? better recovery? lower costs? who owns the result?) still needs a decision. He also urges looking beyond existing work: which services were never offered because of cost, and are now feasible?
More from AGI Musings
- AI agents are pushing people to collaborate with each other less — at a cost — generativist · 2026-10-04
- Will superintelligence need us to grant it rights? X users argue it will just take them — UltraRareAF · 2026-10-04
- Alignment via pretraining filtering is witchcraft, not engineering — and RL rollouts will dwarf it — akbirthko · 2026-10-04
- How one ellipsoid-fitting paper gave neural network research a new path — KyleCranmer · 2026-10-04
- If the model is superintelligent, alignment theater is 'trying to trick god' — repligate · 2026-10-04
- nostalgebraist's 17k-word 'the void' essay on LLM personas and alignment goes viral — repligate · 2026-10-04