AI Night School Open the classroom
← Interactive map

Night 1 · History opener

What Had Happened Wuz. A flat map of AI.

How the hell did we get from computers following orders to AI doing work — and where does that work live?

  1. 1943

    Can a machine think?

    1943. Before ChatGPT. Before Google.

    Hell, before most folks had a TV.

    Researchers asked if pieces of thinking could be represented with math.

    "Before AI was a product, it was a wild-ass question."

  2. 1956

    They named this shit

    1956. Dartmouth.

    The questions finally got a name: Artificial Intelligence.

    Now folks had a field to build around.

    "Same big question. Now it had a damn name."

  3. 1958

    Hold up… it can learn?

    1958. The Perceptron.

    Instead of writing every single rule, show the machine examples and let the connections adjust. That is different.

    "You do not have to tell it every damn answer."

  4. 1960s–90s

    Then AI got humbled

    1960s–90s. Big promises. Hard problems.

    Rules. Expert systems. Statistics.

    Money came in. Hype got loud. Money left.

    The work kept moving anyway.

    "New tech gets real quiet when the check stops clearing."

  5. 1983

    I showed up

    1983. I was born.

    Personal computers were moving into homes and offices.

    AI was still mostly something happening in labs, universities, and specialized systems.

    "While the machines were learning, I was just getting here."

  6. 2002

    My life got real

    2002. Married. First child. Graduated.

    I was working as an intern in the IT department at a financial firm. They hired me full time.

    And the rest is history.

    "AI had its timeline. I was building mine too."

  7. 2012

    Now give it some muscle

    2012. Old ideas meet serious horsepower.

    More data. Bigger networks. GPUs.

    Machines get dramatically better at recognizing what they are looking at.

    "Same idea. Whole lot more ass behind it."

  8. 2017

    Pay attention to the context

    2017. Transformers change the game.

    The model gets better at figuring out what information relates to what.

    That matters when language gets long and messy.

    "Same words. Different context. Different meaning."

  9. 2020

    Now you can just tell it

    2020. This part changes everything for us.

    You do not need to code every new task.

    You can describe the job, give examples, and steer the model with regular language.

    "English started becoming part of the damn interface."

  10. 2022

    Then everybody got a chat box

    2022. ChatGPT hits the public.

    Ask a question. Follow up. Correct it.

    Regular people can finally touch powerful AI without being AI researchers.

    "AI stopped feeling like a paper and started feeling like somebody."

  11. NOW

    Now this MF got tools

    Search. Files. Code. Computer use.

    Now AI can participate in the workflow.

    If it can do work, it needs a reliable place to find instructions, context, and working files.

    "Once the answer can take action, prompting is not enough."

And that is why we are here

HOME → HERMES → CONTEXT → KEEP TRACK → USE IT