AI.info

Fondamenti dell’IA

Explore the AI.info learning paths.

In inglese

• AI literacy basics

  Seventeen lessons that build the judgement to read an AI system, its evidence, and its consequences — before the vocabulary.

• How machines learn

  Twenty-six lessons on what actually changes with experience — from a real problem to a learning task you can defend.

• Kinds of learning

  Supervised, unsupervised, self-supervised: the feedback signal decides everything downstream, including what can go wrong.

• Evaluation

  Turning a number into evidence a decision can rest on: baselines, populations, harms, and what a benchmark cannot answer.

• Mathematical foundations

  The notation an AI system is written in, taught as a modelling language rather than a prerequisite to survive.

• ML data engineering

  The evidence platform underneath every model: prediction time, ownership, contracts, and the leaks that look like accuracy.

• Classical machine learning

  The models that still win on tabular data, and the structured assumptions each one smuggles in.

• Training and optimization

  Objectives, losses and the feedback loop that fits a model — and what a training curve does not tell you.

• Neural networks

  Differentiable programs, traced by hand from a single neuron to an architecture you can defend in a review.

• Deep architectures

  Convolution, recurrence and attention read as one question: how information is routed through a network.

• Unsupervised learning

  Finding structure when there is no answer key, and knowing when the structure is in the data or in the method.

• Advanced techniques

  Ensembles, self-supervision, mixtures of experts — with a decision map for when the simple thing has stopped working.

• Causal inference

  What would have happened otherwise — and how to defend the answer when nobody can observe it.

• Natural language processing

  Text as engineering — corpora, rights, tokens, and evaluation that survives contact with real language.

• Computer vision

  From photons and optics to a measurement somebody has to act on, with the failure modes at every stage.

• Speech and audio

  Sound as measurement — sampling, features, recognition — and the decisions people build on top of it.

• Generative AI

  Generation as a system rather than a magic box: distributions, tokens, grounding, and the controls around the model call.

• Recommender systems

  Decision pipelines that shape what millions of people see, and the objectives that quietly decide for them.

• AI agents

  Action under control: goals, tools, permissions, stop conditions, and who answers for what the system did.

• MLOps

  Production as a control system — contracts, risk tiers, ownership, and what happens at three in the morning.

• Responsible AI

  Governance as sociotechnical work — stakeholders, foreseeable misuse, and a system followed from proposal to retirement.