Since that initial launch, a lot has happened. OpenAI has released the fourth iteration (v. 4.0) of the multimodal (i.e., accepting text and image input) large language model on which ChatGPT is based. Companies including Google and Microsoft (the latter being a major OpenAI partner and investor) have launched competitive and complementary products. In February 2023 Microsoft released a new version of its Bing search engine that includes chat based on GPT-4, effectively turning it into an AI-powered real-time search engine connected to the internet with a conversational interface. Microsoft also launched a separate chat offering called Bing Chat which has the ability to browse the internet. In March 2023 Google launched its own chatbot, Bard, built on a version of the company’s own large language model LaMDA (Language Model for Dialogue Applications). Thus far, ChatGPT appears to be the most accurate of the mainstream offerings.
New competition continues to emerge. In February 2023 Meta announced its own large language model, LLaMA (which was subsequently leaked online), and in April Amazon launched Bedrock, a tool that lets software developers build and scale AI applications. Apple has been relatively quiet, but as the creator of Siri, you can bet they’re not sitting still. Meanwhile, a plethora of smaller competitors have stepped into the spotlight as well.
AI already serves our daily lives in everything from personalized recommendations to search engines, online advertising and navigation, albeit in ways that may not be obvious to end users. However, generative AI’s ability to converse in natural language and create original content in both text and multimedia form opens a vast new world of possibilities
|
