The Generative AI Arms Race: ChatGPT, LaMDA, FAIR, Ernie, and the future of information.
The explosion of interest in ChatGPT and generative AI has caused major tech companies to rethink and revise their product plans. Every tech giant has been investing in AI for years, but the sudden rise of ChatGPT has given the industry a particular interest in generative programs that produce content in response to user prompts. Let’s break down what each major tech company is doing to capitalize on this trend.
Microsoft has had a close relationship with OpenAI, the creator of ChatGPT, and has been investing in the technology for years. As a result, Microsoft is leading the deployment of generative AI-based services. Last week, Microsoft launched a premium version of its Teams groupware that integrates ChatGPT-based tools for summarizing meeting notes, organizing personal tasks, and translating texts. The company is also expected to use ChatGPT-style AI to potentially integrate it with other programs like Word and email.
Additionally, Microsoft is planning to launch a new version of Bing, integrated with OpenAI's ChatGPT technology, which will change the way we search the web. The new Bing features a larger chat box instead of a search bar, allowing users to search using natural language and even have conversations with the AI. The AI will provide results with cited sources and can adjust its search queries based on the user's requirements. The new Bing still allows traditional keyword-based searches with a page of results. Microsoft has not yet announced the launch, but it is believed to be imminent as a preview version was briefly available before being taken down.
Google may be introducing AI features in Search to help users see the big picture and gain insights. The company is also making it easy for developers to innovate with AI by providing them access to its Generative Language API powered by LaMDA, with more models to follow. Google is committed to developing AI in a bold and responsible way, and follows a set of AI principles which were published in 2018. The company provides education and resources for its researchers, partners with governments and external organizations to develop standards and best practices, and works with communities and experts to make AI safe and useful. Google intends to provide a suite of tools and APIs to make it easy for others to build innovative applications with AI, and is excited to help scale these efforts through its Google Cloud partnerships with Cohere, C3.ai, and Anthropic.
Google has been working on large language model-based generative AI for years and was a pioneer in the technology. However, the company has been cautious in releasing its research due to concerns over accuracy, trustworthiness, and potential for bias. The rise of ChatGPT reportedly led Google CEO Sundar Pichai to declare a "code red" and fast-track a number of generative AI projects. In late January, Google demonstrated a new generative AI project that can generate music from a user's text prompt. The company is committed to being bold with innovation and responsible in its approach to AI and will continue to make advancements in this field in the future.
Meta, the parent company of Facebook, has invested heavily in AI and has detailed efforts in the generative space, including text-to-image and text-to-video. Mark Zuckerberg, CEO of Meta, considers AI the foundation of Meta's discovery engine and ads business and is committed to becoming a leader in generative AI. However, two recent public demos of generative AI projects by Meta went poorly, with one project drawing criticism for low-quality output and embracing conspiracy theories, and the other being criticized for inaccuracies. AI and generative AI were mentioned more times on Meta's earnings call than the "metaverse," the emerging technology CEO Mark Zuckerberg has bet the company on. Experts see potential for Meta to use AI to generate and populate virtual worlds, but the company first needs to provide reasons for users to visit them.
Yann LeCun, Chief Artificial Intelligence Scientist at Meta, criticized ChatGPT in a talk hosted by the Collective[i] Forecast. LeCun said that ChatGPT is not innovative in terms of techniques, and it is well put together but not revolutionary compared to other systems built by other companies. He explained that ChatGPT uses Transformer architectures that are pre-trained in a self-supervised manner, a technique that he has been advocating for a long time. LeCun believes that FAIR will eventually catch up to OpenAI and be known for breakthroughs in AI, including generative art and creation aids. Paul Buchheit, the creator of Gmail, disagrees with LeCun and believes that ChatGPT-like AI will destroy Google's business in two years.
Amazon uses AI for various purposes, including Alexa's voice recognition and warehouse optimization, but has not yet entered the generative AI market. Reports of software developers inside Amazon using ChatGPT as a coding aid led a lawyer to order employees not to share corporate information with the bot out of fear that company secrets could leak. Amazon CTO Werner Vogels has stated that ChatGPT "is not concerned about the truth, but just about putting words together convincingly."
Apple CEO Tim Cook spoke about the potential of AI to change the company during an earnings call last week, but the examples he offered were not related to generative AI. Apple remains a secretive company, and if it has a generative AI product in the works, it is unlikely to be revealed until it is ready to launch.
Baidu, less known in the US, is China's largest search engine. It’s said that they will launch their own ChatGPT in March. In late 2021, Baidu released Ernie 3.0 Titan, a language model with 260 billion-parameters. Now, they're using Ernie to power its own ChatGPT-like application.
In conclusion, the trend of generative AI driven by ChatGPT has caused major tech companies to reevaluate their product plans and invest in the technology. From Microsoft leading the deployment of generative AI-based services to Amazon's cautious approach, each company is taking its own path to capitalize on this trend. While AI has a bright future, the future uses and potential harms of generative AI are still unknown.