OpenAI, the maker of the Artificial Intelligence (AI) chatbot, ChatGPT, has launched new AI models, o1-preview and o1-mini, that are capable of performing some human-like reasoning tasks.
These models are specifically designed to tackle complex tasks and solve some challenging problems by mimicking the thought process of humans, taking more time to work through difficult tasks than previous models. According to OpenAI, the o1 thinks before it answers and can produce a long internal chain of thought before responding to the user.
Announcing the launch of the AI models, the company wrote,
Tekedia Mini-MBA edition 15 (Sept 9 – Dec 7, 2024) has started registrations; register today for early bird discounts.
Tekedia AI in Business Masterclass opens registrations here.
Join Tekedia Capital Syndicate and invest in Africa’s finest startups here.
“We’ve developed a new series of AI models designed to spend more time thinking before they respond. They can reason through complex tasks and solve harder problems than previous models in science, coding, and math. Today, we are releasing the first of this series in ChatGPT and our API. This is a preview and we expect regular updates and improvements. Alongside this release, we’re also including evaluations for the next update, currently in development.”
According to OpenAI, these enhanced reasoning capabilities can be particularly useful if users are tackling complex problems in science, coding, math, and similar fields. For example, o1 can be used by healthcare researchers to annotate cell sequencing data, by physicists to generate complicated mathematical formulas needed for quantum optics, and by developers in all fields to build and execute multi-step workflows.
It disclosed that the o1 ranks in the 89th percentile on competitive programming questions, places among the top 500 students in the US in a qualifier for the USA Math Olympiad, and exceeds human PhD-level accuracy on a benchmark of physics, biology, and chemistry problems. In a qualifying exam for the International Mathematics Olympiad (IMO), GPT-4o correctly solved only 13% of problems, while the reasoning model scored 83%. Their coding abilities were evaluated in contests and reached the 89th percentile in Codeforces competitions.
Notably, OpenAI further disclosed that in tests, the next model update performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology. It also excels in math and coding.
OpenAI however says it’s taking a slow and cautious approach to releasing the new models. It’s releasing a couple of “early previews” of two of the models in the series. People with ChatGPT Plus or Teams accounts can access “o1-preview” by choosing it in a drop-down menu within the chatbot. They can also choose “o1-mini,” which is faster and good at STEM questions, OpenAl says.
The launch of the o1 series represents a significant step forward in Al reasoning capabilities, and while it may not yet be the ultimate solution, it is undoubtedly a major advancement in the field of artificial intelligence.
LinkedIn Summary
OpenAI’s newest model, code-named Strawberry — and said to be capable of more complex reasoning — is already ripe for the picking. The newly released model, OpenAI o1, can work through more complicated math, science and coding challenges than OpenAI’s previous offerings, and it can even tackle subjective topics such as product marketing strategies. OpenAI’s business is already booming, with monthly sales revenue tripling since last year, but it’s eager to stay ahead of competitors and keep the money coming in. The company has already demonstrated Strawberry to national security officials.
- OpenAI is talking with investors to raise $6.5 billion at a valuation of $150 billion.