Primus × Cournot’s collaboration: I’m not knocking the technology itself; what I’m knocking is the timing.
I took a look at the discussion on X, and most of it was replies tagged with Kaito: "Very massive collaboration""seems very promising now""revolutionizing onchain finance"… Ten different ways of saying the same thing, with views ranging from single digits to a few dozen. The most-viewed original post got 2,115 views and 32 replies. If you know, you know: that’s yap points at work, not people who actually understand what’s going on.
Now, to the point. zkTLS can prove that this data really came from a particular website or API, and that no one tampered with it along the way. That’s all. It can’t prove that the website itself is telling the truth. If the source reported something wrong, what you get is "cryptographically authenticated incorrect data."
The real arguments in prediction markets aren’t about whether "the data was scraped from the official site." They’re about how to interpret the rules: what counts as "happening," which point in time to use, and whose judgment to follow when the wording is ambiguous. Cournot says it uses AI to reason over verified evidence and produce auditable outcomes. That raises the question: if the AI gets it wrong, who takes responsibility, and who has the authority to overturn its decision? Even their own summary of X news says this system hasn’t fully resolved accuracy issues or disputes over rules.
Breaking down data sources and using AI to interpret them is a clean approach. But no one can even answer how many real users are testing it right now (someone asked exactly that below, and got no reply). Watch the product, not the replies.
#zkTLS #prediction markets
I took a look at the discussion on X, and most of it was replies tagged with Kaito: "Very massive collaboration""seems very promising now""revolutionizing onchain finance"… Ten different ways of saying the same thing, with views ranging from single digits to a few dozen. The most-viewed original post got 2,115 views and 32 replies. If you know, you know: that’s yap points at work, not people who actually understand what’s going on.
Now, to the point. zkTLS can prove that this data really came from a particular website or API, and that no one tampered with it along the way. That’s all. It can’t prove that the website itself is telling the truth. If the source reported something wrong, what you get is "cryptographically authenticated incorrect data."
The real arguments in prediction markets aren’t about whether "the data was scraped from the official site." They’re about how to interpret the rules: what counts as "happening," which point in time to use, and whose judgment to follow when the wording is ambiguous. Cournot says it uses AI to reason over verified evidence and produce auditable outcomes. That raises the question: if the AI gets it wrong, who takes responsibility, and who has the authority to overturn its decision? Even their own summary of X news says this system hasn’t fully resolved accuracy issues or disputes over rules.
Breaking down data sources and using AI to interpret them is a clean approach. But no one can even answer how many real users are testing it right now (someone asked exactly that below, and got no reply). Watch the product, not the replies.
#zkTLS #prediction markets