What is ChatGPT?
OpenAI’s ChatGPT produces text-based answers to prompts. It’s an example of generative AI, a kind of artificial intelligence technology that’s the latest craze in Silicon Valley.
Products like ChatGPT are built on top of generative AI software capable of producing stunning imagery that can look like paintings sketched by real humans or, as is the case with ChatGPT, essays that read like they were written by actual college students.
Investors are pouring billions of dollars into startups that specialize in generative AI. Microsoft, for example, announced a new multibillion-dollar investment in OpenAI in late January. Meanwhile, Khosla Ventures, Craft Ventures, Sequoia, Entrepreneur First and Lux Capital are the top venture capital firms investing in generative AI startups, according to deal-tracking firm PitchBook.
Stability AI, another popular generative AI-focused startup, said in October that it received $101 million in funding from Coatue, Lightspeed Venture Partners, and O’Shaughnessy Ventures. Stability AI built an open-source project called Stable Diffusion that garnered a lot of attention from people who were captivated by its ability to dream up images based on written prompts. — Jonathan Vanian
Bank of America says to expect ChatGPT integration into Bing search
Bank of America analysts said Microsoft’s event will likely focus on ChatGPT’s integration into Bing search, as well as the company’s broader partnership with OpenAI.
“The AI race is clearly on for the tech sector,” the analysts wrote in a Monday note. They added that speed will be important for scale as the AI models are learning.
There’s already competition brewing between Microsoft and Google on the AI front. Google will hold an event on Wednesday to talk about its new Bard AI product. Google announced Bard on Monday and said it will roll out in Google Search in the coming weeks.
— Ashley Capoot
OpenAI’s ChatGPT might pose data-security risks
OpenAI’s buzzy ChatGPT service knows a lot about the world after being trained on loads of online data. Big technology companies are concerned about the risks that it poses around data security.
That reportedly includes Microsoft, which supplies cloud-computing services to OpenAI in order to run ChatGPT.
In January, a Microsoft engineer said in an internal online discussion that Microsoft employees should not tell ChatGPT any sensitive corporate information, according to Insider, which reviewed the warning. OpenAI could use the information in the course of training future models, the engineer wrote. From there, it could theoretically be possible for someone to receive confidential information while conversing with a version of ChatGPT that’s relying on a more up-to-date language model.
Insider also reported that an Amazon attorney advised staff members not to send ChatGPT any confidential information, including source code. — Jordan Novet
People can’t get enough of ChatGPT
OpenAI logo seen on screen with ChatGPT website displayed on mobile seen in this illustration in Brussels, Belgium, on December 12, 2022.
Jonathan Raa | Nurphoto | Getty Images
Since the hybrid AI research firm OpenAI released ChatGPT in November, people can’t seem to get enough of the chat-generating software.
OpenAI CEO Sam Altman said via Twitter that, within 5 days of debuting ChatGPT — which is still in a so-called beta, or experimental version — the software “crossed 1 million users!”
People have found a number of users for the tool, from helping technologists organize research notes for lectures to generating software code on behalf of developers. ChatGPT has become so popular that some schools across the nation have banned students from using the tool as a new “homework assistant.”
Check out this CNBC documentary on the rise of ChatGPT and its potential to shake up the business world. — Jonathan Vanian