All paths

馃

How My Thinking on AI Evolved

Track how my views on AI shifted across the years, what I got right, what I missed.

Reading these in order shows how my framing of AI changed as the technology and ecosystem evolved. From early scepticism to nuanced enthusiasm to calibrated realism.

7 articles
  1. STEP 1路Jul 2017

    AlphaGo Retires and What It Actually Proved

    DeepMind announced AlphaGo would play no more competitive matches after defeating the world number one. The interesting question was never whether it could win.

  2. STEP 2路Jun 2018

    Google Duplex and the Uncanny Valley of AI Voice

    At Google I/O, a recording played of an AI making a haircut appointment by phone. The AI said um and uh convincingly. The reaction was complicated.

  3. STEP 3路Sept 2023

    LLM Agents in 2023: Impressive Demos, Harder Reality

    AutoGPT went viral in April 2023. Agents that could autonomously break down goals into tasks and execute them looked like the next leap. Six months later, the gap between demo and production was clear.

  4. STEP 4路Aug 2024

    AI Engineering: The New Discipline Nobody Had a Name For

    Something new emerged in 2024 that sat between software engineering and machine learning but was not quite either. Building reliable products on top of language models required skills and patterns that had not existed before.

  5. STEP 5路Feb 2025

    Agentic AI in Production: What Works in 2025

    Two years after AutoGPT made agents seem imminent, the production reality is clearer. Narrow agents with human oversight work. Fully autonomous agents are still research. Here is the honest picture.

  6. STEP 6路Oct 2025

    Open Source AI Closes the Gap

    The argument for open source AI used to be mostly about control and cost. By late 2025, capability had joined that list.

  7. STEP 7路Dec 2025

    What 2025 Got Right and Wrong About AI

    Predictions made at the start of a year rarely survive contact with events. Reviewing what the forecasts said and what actually happened teaches more than the forecasts themselves.

End of path

Browse other paths