IntermediateLesson 1 of 59 min read
Who's speaking, and what they're selling
A course from the Academy: Reading AI news without getting fooled
Go to the quiz ↓Before judging any claim about AI, one question sorts things out faster than all the others: who is making it, and what do they stand to gain? That's not cynicism, it's the basics of being a reader.
The five voices, and their stakes
- A lab talking about its own model. It's raising funds, hiring, defending market share. Its announcements are both sincere and slanted.
- An investor already committed. They benefit from the field looking vast. They have no interest in it looking mature.
- A competitor commenting. Their sudden caution about risks is best read in light of their own lag.
- An independent researcher. Generally the most reliable, and the most nuanced, which is why they get quoted the least.
- Someone whose job is under threat. Their concern is legitimate and informed; it is also a position.
None of these voices necessarily lies. All of them look from somewhere, and that somewhere shows in what they choose to say, more than in what they assert.
The most confusing case: the insider warning
A lab executive declaring their own product potentially dangerous seems to act against their interest. That's what makes the statement credible, and it's also what makes it effective.
Saying "what we're building could change the world, and that's why we need regulation" asserts three things at once: that the technology is powerful, that the speaker is responsible, and that the rules should be written with them. It can be perfectly sincere. It is also, in the same breath, an excellent business argument.
Above all, remember the direction of the statement: if this position is followed through, does it advantage or disadvantage the person making it?
What this looks like in practice
You don't have to decide who's right. You have to spot what isn't said: the lab publishing a score doesn't say where it lost, the investor talking growth doesn't talk profitability, the competitor raising alarms doesn't mention their own timeline.
Do this right now: take the last AI announcement you read. Who was making it? What does that person gain if you believe them?
Complete the lesson
A 4-question quiz: 3 correct answers are enough, plus an exercise to practise. It opens in the Academy, on this same lesson.
Take the quiz and complete the lesson Free. With an account, your progress is saved and leads to the course certificate.