TALAMANA · A READY-MADE PATH
The cutting edge — agents
21 ideas, in order. about 35 minutes of listening. It ends at "Agents, goals, tools, memory".
The path
- Machines learn from examples — The examples are the teacher, so bad examples teach badly.
- Prediction, not understanding — The machine produces what is likely, and likely is not the same as true.
- The name is from 1956 — The name was written down in 1955, and the field dates from the 1956 workshop.
- Promises, then winters — A student who knows the winters can tell a demonstration from a delivery.
- Rules that draw: the kolam — Rules can make form, and you have watched it happen before breakfast.
- Early algorithms: the Sulba Sutras — A rule that works is knowledge, even when nobody has written why.
- Scale, ratio, and proportion — A door two metres wide is drawn two centimetres wide, and every length does the same.
- Talamana: proportion as a rule system — A proportion rule is somebody's judgment written down so it can be repeated.
- Chance, and how to name it — A number between 0 and 1 names chance, and machines run on it.
- Same input, same output? — A rule is checked by reading it, and a draw by looking.
- Learned, not programmed — Nobody wrote this behaviour line by line, so there is no rulebook to read.
- 2012: the examples won — The ideas were decades old, and enough labelled data and cheap computing arrived around 2012.
- Attention becomes the architecture — Most chatbots and many image tools share one 2017 ancestor, and so share its limits.
- Language models, the next-word engine — Nothing in the next-word loop checks whether the argument is true.
- Tool use, reaching the world — The model proposes a tool call and the runtime executes it, so permissions live there.
- Everything it knows is data — To the machine a picture, a word and a song are all numbers that people collected.
- Small models, large models — Size is one side of a trade, and size alone does not rank models.
- What "model" really means — A model is a trained mathematical function, and the manners come from the app.
- Training and inference are different times — Training set the weights before you opened the app, and your chat never changes them.
- Weights, context, retrieval, memory — Only the weights are the model, and the other three are added around it.
- Agents, goals, tools, memory — An agent acts over several steps, so decide in advance where the human gate sits.
Take this path on the map · The complete map · Logika · RBDS AI Lab, India.