that of The New York Times, is not exactly easy. In fact, on some days the solution rate drops below 50 percent. To show where players are making mistakes, a new bot uses AI to identify the most common mistakes each day.

For the uninitiated: Connections is a word/logic game that the newspaper launched last year. Every day you are presented with a grid of 16 words that you have to sort into four categories. There is only one solution and after four mistakes the game is over. However, it is not that easy. There are often lots of red herrings and often at least five possible answers for a group.

I’m addicted and have been playing every day since last July. It’s one of our (though it may be ).

After you have won or lost the game of a day, you can stroll to the. Like the bot for Wordleyou’ll see how well you did compared to other players and receive a skill score out of 99. This is based primarily on how few mistakes you make, but you’ll get extra points if you solve the harder purple and blue categories first.

After seeing the skill score and other details (like whether you got a red herring), the AI ​​feature comes into play. This highlights the most common incorrect guesses of the day. It also tries to guess a description of the group that players had in mind. So if you don’t guess Gutter, Bowl, Alley, and Lane, the bot might think you were looking for a list of bowling-related terms. This is a real example from a recent game where I made this exact mistake. Alley and Lane were actually street types.

However, your own misjudgments may not be reflected in the bot. This is because there are about 2.6 million different ways to summarize each grid. Keep in mind that while you cannot Just Account to play ConnectionsYou must be logged in to one to use the bot and track your results.

Another interesting thing about the bot is that it is the first time that the Just’ The editorial team will regularly publish AI-generated English texts. Before publication, the newspaper’s editors will edit them for style and clarity. This means that the Just points out that “there is no way to use math or reliably solve the game,” so you can’t really use programs like ChatGPT to cheat yet.