Sam Altman, CEO of OpenAI, at an event in Seoul, South Korea, on June 9, 2023.
Bloomberg | Bloomberg | Getty Images
OpenAI announced its recent, more powerful GPT-4 Turbo artificial intelligence model Monday during its first in-person event, and revealed a recent option that can let users create custom versions of its viral ChatGPT chatbot. It is also cutting prices on the fees that firms and developers pay to run its software.
OpenAI’s announcements show that one in all the most popular firms in tech is rapidly evolving its offerings in an effort to remain ahead of rivals like Anthropic, Google and Meta within the AI arms race. ChatGPT, which broke records because the fastest-growing consumer app in history months after its launch, now has about 100 million weekly lively users, OpenAI said Monday. Greater than 92% of Fortune 500 firms use the platform, up from 80% in August.
The event also included a surprise appearance by Microsoft CEO Satya Nadella.
“The systems which might be needed as you aggressively push forward in your road map require us to be on the highest of our game, and we intend fully to commit ourselves fully to creating sure you all… haven’t only the perfect systems for training and inference, but in addition essentially the most compute,” Nadella told OpenAI CEO Sam Altman while onstage together. He added, “That is the way in which we’ll make progress.”
Earlier this yr, Microsoft’s expanded investment in OpenAI — an additional $10 billion — made it the largest AI investment of the yr, in line with PitchBook, and in April, the startup reportedly closed a $300 million share sale at a valuation between $27 billion and $29 billion, with investments from firms corresponding to Sequoia Capital and Andreessen Horowitz. As recently as last month, OpenAI was reportedly in talks to shut a deal that might result in an $80 billion valuation.
In his speech Monday, Altman said the day’s announcements got here from conversations with developers about their needs over the past yr.
Here’s what OpenAI announced Monday:
GPT-4 Turbo is the newest AI model, and it now provides answers with context as much as April 2023. Prior versions were cut off at January 2022. For instance, for those who asked GPT-4 who won the Super Bowl in February 2022, it would not have been in a position to let you know. GPT-4 Turbo can.
“We are only as annoyed as all of you, probably more, that GPT’s knowledge concerning the world resulted in 2021,” Altman said in a speech Monday.
It also accepts lots more input. While earlier versions limited you to about 3,000 words, the GPT-4 Turbo accepts inputs of as much as 300 pages in length. It means you might ask it to summarize entire books.
GPT-4 also supports DALL-E 3 AI-generated images and text-to-speech. It also has six preset voices to select from, so you possibly can select to listen to the reply to a question in a wide range of different voices.
OpenAI said GPT-4 Turbo is offered in preview for developers now and will probably be released to all in the approaching weeks.
OpenAI said it is also cutting the costs for developers. “GPT-4 Turbo input tokens are 3x cheaper than GPT-4 at $0.01 and output tokens are 2x cheaper at $0.03,” the corporate said, which implies firms and developers should save more when running a number of information through the AI models.
Personalized chatbot builder
Until now, ChatGPT’s enterprise and business offerings were the one way people could upload their very own data to coach and customize the chatbot for particular industries and use cases. Now it’s adding the choice for anyone to create custom chatbots.
AI “agents” are one in all the buzziest uses of the technology recently, with many startups vying to supply the type of personalized AI tools that buyers may already be acquainted with via popular culture representations, corresponding to Tony Stark’s J.A.R.V.I.S. in Marvel movies, or Pam in Disney Channel’s Smart House.
“Anyone can easily construct their very own GPT—no coding is required,” the corporate wrote in a release. “You’ll be able to make them for yourself, just on your company’s internal use, or for everybody. Creating one is as easy as starting a conversation, giving it instructions and additional knowledge, and picking what it may possibly do, like searching the net, making images or analyzing data.”
Greater than two million developers constructing their very own tools using ChatGPT’s API may also have the ability to customize the chatbot, meaning consumers will likely see personalized AI chatbots popping up in lots of more places, including apps and web sites they use commonly.
Now that users and developers can launch their very own, personalized AI chatbots, OpenAI is introducing a recent revenue driver for the corporate: Its own version of the app store.
The GPT Store allows individuals who create their very own GPTs to make them available for public download, and in the approaching months, OpenAI said people will have the ability to earn money based on their creation’s usage numbers.
“Once in the shop, GPTs change into searchable and will climb the leaderboards,” the corporate wrote in a release. “We may also highlight essentially the most useful and pleasant GPTs we come across in categories like productivity, education, and ‘only for fun.'”
Until Monday, ChatGPT users needed to hop between different apps and web sites to make use of all of OpenAI’s tools, which contributed to a rather higher learning curve. On Monday, the corporate announced it has streamlined its AI tools into one place: Using ChatGPT now offers image generation via DALL-E, browsing, data evaluation, document upload and PDF search. Before now, Anthropic’s Claude was the one competitor chatbot to permit PDF search.
As generative AI-related legal motion is on the rise, Altman announced Monday that OpenAI will “step in and defend our customers” and “pay the prices incurred for those who face legal claims around copyright infringement.” It echoes similar statements made by Google, Microsoft and Adobe.