OpenAI launches 'reasoning' model

Yesterday, OpenAI released its latest model, O1, which marks a significant leap from chatbots to sophisticated reasoners. Named to indicate that it is resetting ‘the counter back to 1’, O1 pushes the industry beyond conversational AI to models capable of solving complex, multistep problems with near-human-level reasoning. ?? Reason: O1 excels at complex tasks. On a qualifying exam for the International Mathematics Olympiad, it correctly solved 83% of problems compared to GPT-4o’s 13%. In Codeforces programming contests, it ranks in the 89th percentile, with claims it will soon rival PhD students in physics, chemistry, and biology. ?? Expensive: Quality comes at a cost. At 3-4x the cost of GPT-4o, O1-preview comes with a hefty price tag—$15 per million input tokens (vs $5), $60 per million output tokens (vs $15) ?? Slow: O1 mirrors human problem-solving, taking extra time to think through responses. This deliberate "thinking time" may slow performance but significantly enhances accuracy. ?? Limitations: O1 struggles with factual knowledge, lacks browsing capabilities, and can’t process images or files. While it reduces hallucinations, they’re not entirely gone. The cost-performance trade-off is real: not all tasks will need O1’s advanced reasoning, and its higher cost and latency might limit its applicability. But for use cases demanding deeper problem-solving, the potential is enormous. For those, the question isn’t how quickly AI can respond—it’s how deeply it can think. As one OpenAI researcher noted, O1 thinks in seconds—future models might think in hours, days, or even weeks, solving world-changing problems like curing cancer or innovating new battery technologies. It’s clear the future of AI isn’t about one model to rule them all. To each use-case, its own solution!

OpenAI's Latest Leap: The "Strawberry" Reasoning Model ?? OpenAI has done it again with its new, more powerful reasoning model, code-named Strawberry, officially branded as OpenAI o1. This model can tackle more complex challenges than its predecessors, excelling in math, science, and coding, while even demonstrating prowess in subjective topics like product marketing strategies. ?? Why It Matters for Business & Tech: The advances in AI reasoning models signal a major shift in how we tackle everyday tasks, from financial planning to coding, and even strategic thinking. AI is no longer just about automating repetitive tasks but elevating the decision-making process itself. For investors, this surge in OpenAI’s capabilities underscores huge growth potential, especially as AI integrates deeper into various industries. ?? What’s Next for OpenAI? OpenAI has bigger plans on the horizon. They're in the development phase for a new flagship large language model called Orion, which promises to push the boundaries of reasoning, logic, and creativity even further. They’ve already demonstrated Strawberry to national security officials, underlining its potential in mission-critical applications.The business case for this model is clear — OpenAI’s monthly sales have tripled compared to last year, and they are now looking to raise $6.5 billion at a staggering valuation of $150 billion. ?? My two cents AI isn't replacing professionals — it's enhancing what we can do. We should spend time training and improving the way we do everyday tasks, as well as strategic thinking. ?It takes time, but it’s worth it. Are you ready to leverage this next wave of innovation? Check more info here: https://lnkd.in/eewD3w2n #AI #Strawberry

If AI had a flavour, this week it would ?????????? ???????? ???????????????????????? ??. OpenAI ’s latest LLM model, o1 (known before as Strawberry) is giving #investors a sugar rush. The release happens at a time when OpenAI is raising money at a valuation of $150 billion. Pre-money. ?? The market dislocation between AI #business fundamentals and expectations is increasing Do the business fundamentals justify such a valuation? Does it matter? Let's look at some commercial pointers. ?? ?????? ???????????? ?????? ?????????????????????????? Last month, Meta dropped a bomb on the GenAI industry by offering more than OpenAI does and for free. Meta is making its flagship model, Llama 3.1. 405B, available to anyone who wants it. This version can do pretty much anything that OpenAI, Gemini, or Anthropic are capable of, and just as well (reportedly). Moreover, companies may do so at a fraction of the cost because of its smaller weight. ?? ?????????????????? ???????????? Building commercial use cases still requires expensive computing. Meta timed the release with Nvidia's AI Foundry launch. Llama 3.1 can be customised, packaged, and run on Foundry AI from day one. This ecosystem should attract developers. Overnight, Meta and Nvidia have undercut a whole industry by smashing the entry barriers to compete in this space. ?? ???????????? ????????: ?????????? ???? ?????? ???????????? ????????? This is bad news for any company hoping to sell something Meta has just given away for free. The race for the bottom in GenAI pricing has started. The picture gets messier when one considers the financial situation of OpenAI. OpenAI will burn $5 billion this year ($3.5 billion in projected revenues, minus 7 billion compute spend and $1.5 billion staff costs). Which takes me back to the OpenAI valuation headline: ?How will they justify revenue forecasts? ?Even if o1 lives up to the expectations (i.e. correctly reason logical conclusions from first principle), will that become a commercial blockbuster? ?Where are the viable use cases? (Show me the user. I show you the money) Meanwhile, like in Harry Styles's song, the market ?????????? ???????? ??????????????; ?????? ???????????? ???? ????????. TGIF ?? Happy weekend! i-Expand #commercialization #biotechology #digitalhealth #healthtech #healthcare #uae #venturecapital #entrepreneurship

??OpenAI, the maker of ChatGPT, has launched its highly anticipated "o1" model family (also known as strawberry.) These new “reasoning” models are designed to “think hard about problems before answering” and handle more complex tasks, achieving PhD-level performance in fields like science, healthcare, and technology. This also means that the AI will get a lot better at following your instructions and the quality of outputs will improve in many use cases. The "o1" models are available to ChatGPT Plus and Team users, with usage limited to 30-50 messages per week. But OpenAI notes that these models currently lack some key features of ChatGPT like web browsing and file or image uploads, and that GPT-4o may still be better for many common use cases. -- ??If you want to keep up to date on the most essential AI developments, insights, news and tools, sign up for my weekly AI newsletter (Link in my bio and in the comments ??)

Well, it's here and OpenAI's o1preview (rumored as Strawberry) looks pretty good. I'm going to spend some time diving into it today... Overall, the big innovation here is that the model takes time to think about the answer to your question before diving in and it's able to *greatly* improve on it's results this way. OpenAI has published quite a few videos and posts about it and I suggest diving in to learn a whole lot more here. https://lnkd.in/gfqbHaKy #openai #o1preview #machinelearning #llms #strawberry

Open AI just released a new model called o1 This model is not as capable as the GPT series, but it seems to excel in some fields. So, what should you use it for? Here are some ideas: - Education and Tutoring: With its strong performance in math and science, OpenAI o1 could be an excellent resource for students looking for help with homework or preparing for exams, particularly in STEM subjects. It could act as a virtual tutor, providing detailed explanations, working through problems step by step, and offering alternative solutions or methods. - Research and Development: In fields like physics, chemistry, and biology, where it has shown to perform at a level comparable to PhD students, OpenAI o1 could assist researchers by generating hypotheses, designing experiments, or interpreting complex data sets. It could also be used for simulating experiments or predicting outcomes, speeding up the research process. - Software Development and Debugging: Given its coding abilities and performance in Codeforces competitions, OpenAI o1 could be a valuable tool for software developers. It could help in writing, reviewing, and debugging code, suggesting optimizations, or even generating code snippets based on requirements. This could improve development efficiency and code quality. - Problem Solving and Decision Making: For complex reasoning tasks, OpenAI o1 could be used in various industries to optimize processes, solve logistical challenges, or make data-driven decisions. Its ability to think through problems and refine its reasoning could be beneficial in strategic planning, financial modeling, or operational optimization. - Mathematical Modeling and Simulation: With its strong performance in math, OpenAI o1 could be used for creating mathematical models, solving complex equations, and performing simulations that are critical in fields such as engineering, economics, and environmental science. - Competitive Programming and Education: Given its high percentile performance in Codeforces competitions, OpenAI o1 could be used to train and prepare individuals for competitive programming contests. It could provide insights into problem-solving strategies, algorithm optimization, and coding best practices. Have you tried it yet? It's available starting today.

OpenAI just dropped a new AI model and it's called OpenAI 01. This thing is wild. Check out this video of it coding an entire video game from just a prompt. Now, OpenAI 01 is designed to think harder and solve way more complex problems than we've ever seen before. First off, this AI model is crushing it in math and programming. In a qualifying exam for the International Mathematics Olympiad, 01 correctly solved 83 percent of the problems. For comparison, the previous model GPT 40 only got 13%. That's a massive leap. But wait, there's more! In competitive programming contests, O1 ranked in the 89th percentile. That means it's outperforming most human coders. Now, here's the really cool part. The longer O1 thinks, the better it does on reasoning tasks. It's almost like it's mimicking how we humans tackle tough problems. We don't just blurt out the first answer that comes to mind, right? We think it through, try different approaches. And sometimes even recognize when we're making mistakes. That's exactly what O1 is doing, but at superhuman speeds. OpenAI trained this model to refine its thinking process, just like a student learning to solve complex problems. And get this, in tests on challenging tasks in physics, chemistry, and biology, O1 performs similarly to PhD students. Well, if you're a ChatGPT Plus user, you can access O1 right now. I can't wait to see what people do with this new AI. I can't wait to see what people do with this new AI. Maybe we Will see breakthroughs in scientific research.? Revolutionary new software? The possibilities are mind-boggling!

OpenAI releases "Strawberry" -01 preview model focused on reasoning by incorporating an Ai agent into multi-step complex task execution. Great article from Ethan Mollick with experiments and results. He writes: "Using o1-preview means confronting a paradigm change in AI. Planning is a form of agency, where the AI arrives at conclusions about how to solve a problem on its own, without our help." Incorporating Ai agents is an inevitable step in the evolution of LLMs and Generative Ai. We will indeed be facing some interesting conversations on the future of work as Ai continues to take on more and more complex activities. #Ai #artificialintelligence #generativeai #genai #chatgpt #strawberry #reasoning #futureofwork https://lnkd.in/gcFrinis

?? Today marks an incredible milestone as we release the new OpenAI o1 series of AI models designed to think deeper before they respond. This is one of the moments that truly reminds me why I joined OpenAI. We’re launching OpenAI o1-preview and OpenAI o1-mini, models that represent a breakthrough in reasoning capabilities. These models don’t just mimic human thinking—they explore different strategies, self-correct, and reach PhD-level performance across disciplines like physics, chemistry, biology, and beyond. For the public sector, the potential is immense. Imagine the ability to tackle complex governmental challenges with AI models that can deeply reason and adapt—whether it’s for scientific research, operational efficiency, or policy decision-making. Excited to see how these advancements will reshape public sector innovation! #AI #PublicSector #Innovation #OpenAI