On Nov. 30, 2022, a then-unknown start-up called OpenAI sent shockwaves around the world. OpenAI is the developer of ChatGPT, a model that uses generative AI to answer queries, write software code, generate images, and much more.
Almost instantly, big tech companies began plowing billions into alternative AI model developers to compete with OpenAI.
E-commerce and cloud computing behemoth Amazon (AMZN 2.24%) struck a partnership with Anthropic, pouring $8 billion into the start-up over the last two years. While OpenAI may have the first-mover advantage, a new update from Amazon and Anthropic showcase just how serious the competition is becoming.
Did Amazon just say checkmate to OpenAI? Read on to find out.
ChatGPT and competing platforms such as Amazon's Claude are known as large language models (LLMs). LLMs are pretty easy to use: Similar to searching on Google, all you need to do is type a question into the model's browser and within seconds you will be given a detailed response.
Amazon just took this concept to the next level by releasing something called a hybrid reasoning model. Essentially, hybrid reasoning models come in two forms: standard or extended thinking.
In standard mode, you can ask Claude a question and receive as detailed or as simple an explanation as you'd like. In extended thinking mode, however, the model leverages reasoning applications to provide the end user with a "'chain of thought'" before generating an answer to a question.
By integrating this step-by-step approach to answer a question, users can learn how Claude thinks (or reasons). As such, you can further refine your parameters or dive deeper by asking follow-up questions related to a specific part or step in the model's answer.
Image source: Getty Images.
On the surface, it might look like OpenAI got caught resting on its laurels and allowed the competition to run right by. The fact is that most mainstream LLMs offer a host of overlapping features, as well as contain some differentiating applications.
While Amazon may have beaten OpenAI to the punch in terms of releasing a formal hybrid reasoning model, I wouldn't say ChatGPT's days are numbered. In fact, earlier in February, OpenAI CEO Sam Altman posted a long-form analysis on X (formerly Twitter) in which he admitted that the OpenAI platform has become "complicated" due to various model options presented to users. As a result, OpenAI is working on a way to create "systems that can use all our tools."
To me, that sounds a lot like giving users the flexibility to toggle between a basic, standard model or a more sophisticated reasoning platform -- such as the one just released by Amazon and Anthropic.
While it will take time for OpenAI to get this unification process right, I think Altman's vision suggests that Amazon's new Claude update is not necessarily a checkmate move. Although I see the appeal of a hybrid reasoning model, I do not think enterprises using ChatGPT will suddenly begin shifting to Claude in droves.
Rather, I think the value add of a model capable of standard or extended thinking could ignite some near-term buzz for Amazon and in a best-case scenario potentially lead to a slight jolt for Amazon Web Services (AWS). In the long run, however, I think there will be multiple LLMs capable of hybrid reasoning, and I see OpenAI emerging as a leading player in that field when the time comes.
免責聲明:投資有風險,本文並非投資建議,以上內容不應被視為任何金融產品的購買或出售要約、建議或邀請,作者或其他用戶的任何相關討論、評論或帖子也不應被視為此類內容。本文僅供一般參考,不考慮您的個人投資目標、財務狀況或需求。TTM對信息的準確性和完整性不承擔任何責任或保證,投資者應自行研究並在投資前尋求專業建議。