Skip to content

Topic · Themes

Understanding AI: all our articles

All 122 Alyriaz articles on the “Understanding AI” topic: news, analysis and explainers, newest first. To stay up to date, you can get an alert as soon as a new article on this topic is published.

Filter in the library

  1. · Society

    Anthropic tells its investors it was profitable in the second quarter, with gross margins above 80%

    According to the Financial Times, Anthropic has told its investors it is targeting profitability in the second quarter, with gross margins above 80% on $11.5 billion in revenue. We explain why the margin is the decisive figure.
  2. · Explainers

    Giving an agent access to your accounts, without giving everything away

    Granting an agent access is both necessary and risky. We explain the principles of least privilege as applied to AI agents, and the concrete settings to check.
  3. · Explainers

    Can a model's intelligence be stolen?

    Model extraction involves reconstructing a system's capabilities by querying it at scale. We explain the mechanism, the possible defences, and why none of them really works.
  4. · Society

    Washington accuses Chinese labs of siphoning intelligence from American models

    US agencies accuse Chinese labs of extracting the capabilities of American models at scale. We explain the technique in question and why it is hard to prevent.
  5. · Explainers

    A model that writes a text in one go instead of word by word

    Diffusion models applied to text generate an entire response in parallel rather than word by word. We explain the principle, its speed advantages and its limitations.
  6. · Explainers

    The problems of the millennium, and how a proof comes to be accepted

    Seven Millennium Problems were posed in 2000. We explain what they are, why only one has been solved, and above all how the mathematical community decides that a proof holds up.
  7. · Models

    Qualcomm and Amazon team up on inference chips

    Qualcomm and AWS have announced a collaboration to jointly design chips dedicated to inference. We analyse what this tie-up reveals about the reshaping of the silicon market.
  8. · Explainers

    What is formal verification? Having a machine check a line of reasoning

    Formal verification makes it possible to mechanically check every step of a line of reasoning. We explain the principle, its limits, and why it is becoming central as models produce proofs.
  9. · Explainers

    One image, how many tokens? Why we bill photos by the word

    Image models are billed in tokens, the same unit as text. We explain how an image is split into tokens, why editing costs double, and how to reduce your bill.
  10. · Explainers

    What is an exaflop? Measuring computing power, and why comparisons mislead

    Computing power announcements are multiplying in exaflops. We explain this unit, why the same hardware can show very different figures, and how to read these announcements.
  11. · Explainers

    What model size for which use case?

    Most tasks don't require the most powerful model. We explain which size suits which use case, and why choosing too large costs more without adding anything.
  12. · Explainers

    All the September deadlines: what changes, and when

    Several pricing changes and service closures take effect in September 2026. Here's a round-up of the confirmed dates, what they involve, and what you need to check beforehand.
  13. · Explainers

    How does an AI transcribe speech?

    Speech recognition turns a continuous signal into words. We explain the mechanism, why certain situations remain difficult, and how to get better results.
  14. · Explainers

    What is an agent "harness"? The wrapper that does half the work

    Agent evaluations now measure a model + harness pairing. We explain what this software wrapper is, why it weighs as heavily as the model, and what it changes when choosing.
  15. · Explainers

    What is a model's "effort level"? The setting that changes everything and that nobody looks at

    Recent models offer a reasoning-effort setting that profoundly changes their outputs and cost. Here’s what this parameter does, why comparisons depend on it, and how to tune it.
  16. · Society

    Sony and Warner sue Anthropic over song lyrics, targeting founders personally

    Sony Music Publishing and Warner Chappell have filed a lawsuit against Anthropic and its founders over the use of song lyrics in training. We analyse the strategy and what the books precedent changes.
  17. · Vision

    Can you learn without a teacher?

    AI can explain, rephrase, correct, and adapt to each person's level. We examine what a teacher does that it doesn't, and why that changes the nature of learning.
  18. · Explainers

    What is a CVE vulnerability, and how it gets reported

    CVE, severity score, coordinated disclosure: here's what the security flaw vocabulary means, and what to do in practice when an alert concerns a tool you use.
  19. · Explainers

    Build or buy your AI tool: the decision framework

    Agentic coding tools make internal development more accessible. Here is a practical decision framework for choosing between building and buying an AI tool.
  20. · Explainers

    How does an AI search the web?

    When an AI cites web sources, it follows a four-step chain, each of which can fail. We explain the mechanism and what to check.

Other topics in “Themes”

All the other topics

Every article, month by month →