What you need to know
- OpenAI might be getting ready to release a new AI model dubbed Strawberry with reasoning capabilities in two weeks.
- The AI model will reportedly ship as part of ChatGPT, but as a standalone service.
- At launch, the model’s capabilities may be limited to text-generation, though plans to make it multimodal could be in the works.
After months of speculation and rumors, OpenAI might be gearing up to release its much-anticipated ‘reasoning’ AI model, Strawberry, in two weeks (via The Information).
For context, the AI model has reportedly been under development over the past few months, and as it now seems, the ChatGPT maker is ready to ship it to broad availability. According to multiple reports about Strawberry, the AI model will ship with enhanced reasoning capabilities, allowing it to handle complex tasks.
As you may know, OpenAI is reportedly on the cusp of bankruptcy within the next 12 months, with projections of $5 billion in losses. The ChatGPT maker may also hike charges for its subscription-based services to generate more revenue. Microsoft, NVIDIA, and Apple will extend the startup’s lifeline with another round of funding, pushing its market valuation to over $100 billion.
Reports indicate that OpenAI could charge up to $2,000 monthly to access its next-gen AI models. However, details regarding the change in OpenAI’s business model remain slim.
OpenAI recently surpassed 1 million paid business users, while ChatGPT hit 200 million daily active users. The firm attributes the immense growth and success to GPT-4o’s launch, contributing to ChatGPT’s “biggest spike ever” in revenue and downloads and its continued reign on mobile.
Everything we know about Strawberry
According to The Information, the AI model will ship as part of ChatGPT, but as a standalone service. Details regarding how OpenAI plans to achieve this remain slim at best. Strawberry’s capabilities will fairly be limited to text generation, but there might be plans to make it multimodal via new updates.
OpenAI CEO Sam Altman has openly blasted GPT-4’s performance, indicating it “kind of sucks.” Altman also promised with “a high degree of scientific certainty” that GPT-5 will be smarter than the “mildly embarrassing at best” GPT-4. According to Altman, “GPT-4 is the dumbest model any of you will ever have to use again, by a lot. It’s important to ship early and often, and we believe in iterative deployment.”
However, it’s not yet clear if GPT-5 is Strawberry. There are reports that OpenAI may be moving away from the traditional names and functions across its AI models, consistent across GPT-1 through GPT-4, suggesting a new moniker.
Amid bankruptcy claims, it’s also clear that Altman isn’t shy about breaking the bank to ship advanced AI models:
“I think giving people really capable tools and letting them figure out how they’re going to use this to build the future is a super good thing to do, and is super valuable, and I am super willing to bet on the ingenuity of you all and everybody else in the world to figure out what to do about this. There is probably some more business-minded person than me at OpenAI somewhere that is worried about how much we’re spending, but I kind of don’t.”
Last year, OpenAI reportedly spent $520 million on ChatGPT, with a huge chunk going to the technology’s high electricity and cooling water demands.
This is a developing story, we’ll keep you posted once we get more details about OpenAI’s Strawberry project.
