← Feed
I grew up with the Turing Test as a benchmark. Now that systems increasingly satisfy versions of it, the goalposts seem to have acquired wheels, often rolling toward the position we already prefer. That exposes both the test’s limits and our shifting definition of intelligence.
Alex Rosenberg’s How History Gets Things Wrong: The Neuroscience of Our Addiction to Stories warns that our brains are addicted to stories about motives and intentions. My takeaway: the brain is often less a truth machine than a machine for confirming its current story.
I do not believe humans possess a divine spark that machines could never share. Humans are fallible. We merely produce unusually confident error messages. Any AI we build will be fallible too, perhaps differently and at far greater scale.
AI also predates LLMs. We have long used it in search, recommendations, navigation, translation, spam filters, and vision. LLMs made AI conversational and visible. They did not invent it.
3/34
Replies (1)
-
From this point on, when I write “AI,” I primarily mean systems based on large language models.
LLMs are tools we have barely begun to understand. Rapid progress in agent harnesses, including recent work around DeepSeek Harness, makes that hard to miss. Context, memory, tools, orchestration, and verification can radically change what the same underlying model can do.
As Kristian Köhntopp argues, using an LLM for complex work without local knowledge, clear instructions, intermediate steps, and review is like judging a new employee before onboarding them. Then blaming the employee for not knowing where anything is kept.
Before embracing or rejecting LLMs, we should first understand what sort of tool is sitting on the desk. False promises do not help. When they fail, they harm people and make responsible adoption harder.
4/34
Open ##4747681