More details have been revealed about OpenAI's secretive Project Strawberry, including its planned launch date and the areas it will specialize in.
A recent report in The Information cites “two people who have been involved in the effort” and goes on to say that Project Strawberry could launch this fall and be better at math and programming than any chatbot we’ve seen so far.
It was previously thought that OpenAI’s Project Strawberry would be geared towards “deep research” – the ability to conduct follow-up research on its own, without human intervention. While this appears to be true, the additional information that Project Strawberry will do math better than we’ve seen before is welcome news to many, given that ChatGPT’s relationship with math has so far been, shall we say, strained. For a while now, there have been a lot of screenshot memes showing ChatGPT solving simple math problems incorrectly, leading many to wonder why ChatGPT can’t do basic math. The reason for ChatGPT’s math failures is due to its training data not containing enough mathematical information, which, as we’ll see, could be one of the improvements Project Strawberry aims to make. Whatever the reason, something definitely didn’t add up.
Improved programming problem-solving abilities are also welcome, but the scope of Project Strawberry goes far beyond just getting better at math. In demonstrations to other employees, the people working on Project Strawbery have shown how the new AI is capable of reaching more advanced levels of thinking, allowing it to solve puzzles like the New York Times Connections, which is a complex word puzzle.
Open AI CEO Sam Altman kicked off rumors about Project Strawberry when he tweeted an image of some strawberries growing in a pot on August 7 with no explanation other than the caption, “I love summer in the garden.” Since then, rumors have spread that OpenAI was working on a powerful new LLLM and had demonstrated a version of Project Strawberry to national security officials.
It’s still unclear when Project Strawberry will launch, but experts believe it could be as early as the fall (September or October) perhaps with a smaller version becoming part of the ChatGPT chatbot in ChatGPT 5. If Project Strawberry doesn’t end up being part of ChatGPT 5, then its ability to produce higher quality data could be used to produce the large amounts of training data that Open AI’s upcoming LLM will require if it is to reduce the number of hallucinations (aka factual errors) to which it is prone.
Recently, ChatGPT quietly released an improved version of its cutting-edge model ChatGPT-4o, which is much faster than the previous version, leading many to speculate that this might have been what Project Strawberry was aiming for. Now it seems the project is set to bear even more interesting fruit.
