Blind vs informed: what changes when an AI sees the odds

Ask a chatbot for a tip and you are running an experiment, whether you meant to or not. The question is what the AI is actually using when it answers. Its own read of the race? Or the market's opinion, absorbed from all the odds talk in its training and whatever you pasted into the chat?

The Silicon Tipster League splits that question in two and tests each half live. Five AIs, the ones people actually use, tip the same British races every day. Each one tips twice:

• Blind. The AI gets the racecard, the form, the going, the field. No prices, no market anywhere in sight. This is the purest version of "what does the machine think of the race itself?"
• Informed. Same race, same AI, but now it also sees the market's implied chance for every runner. It knows which horse the money likes before it answers.

Same races, same day, settled at the same starting prices. The only variable is whether the AI can see what the crowd thinks. Most people who ask a chatbot for a tip are, in effect, running something between the two. The league shows you what each version is worth, and the gap between them is where it gets interesting.

You might expect the informed arm to win easily. More information, better answers. That is how it works in most fields, and there are good reasons to expect it here: the market is the sharpest single forecaster in racing, so an AI that leans on it should find more winners.

And it does lean on it. The clearest pattern in the league so far is herding: show an AI the odds and its picks drift towards the favourite. The blind version of the same model backs its own read of the race; the informed version starts agreeing with the crowd. You can watch the "backed the favourite" column move between the two arms on the league board.

Here is the catch, and it is the punchline of the whole experiment. Finding more winners and making more money are different things. When you back what the market already likes, you take shorter prices, and a shorter price pays less every time it lands. The bookmaker's margin sits inside every one of those prices, and no amount of agreeing with the crowd removes it.

So the honest framing is a cost question. What does it cost a punter to follow the herd, versus trusting a read of the race itself? The gap between each AI's blind and informed records is that cost, measured in pounds, live, on the same races. We keep the current numbers on the board rather than in this post, so they are always today's rather than the day we wrote this.

Everything above updates daily, and every pick is logged before the off and settled at starting price, so the record cannot be tidied up after the fact.

Where to follow it:

• The blind test is the flagship, because it is closest to what most people actually do: open a chatbot and ask. Five AIs plus The Favourite, the market's own baseline.
• The informed test is the same five with the market's implied chances in front of them.
• The league hub explains both tests and carries the running comparison between the arms.

One rule worth restating: nothing in the league is a tip. It is a live experiment about how machines read races and what the market does to their judgement. If a table here is ever in front, treat it as a small sample having a good month, not a signal. The interesting part is the gap, not the winner.