ChatGPT, the popular AI chatbot by OpenAI, will launch several new features for paying customers in the coming weeks, allowing users to speak to the app and have the chatbot respond in kind, as well as upload images that can be analyzed by the AI.
The voice features will only be available on the mobile app, while the image analyzing feature will be available on all platforms. Currently, users input text to the chatbot and receive a text response. The new features will allow for a more natural conversation with the bot and the image processing feature will greatly expand its capabilities.
Users could, for example, take a picture of their car engine and have ChatGPT identify the different parts for them, or provide a tutorial on how to check the oil in that particular model.
The AI industry has been in a heated arms race over the past few years, with billions pouring into the industry on a regular basis. Earlier this year, Microsoft invested $10 billion into OpenAI, which was later valued between $27 billion and $29 billion during a $300 million share sale.
Meanwhile, Google has been adding new features to its Bard AI chatbot regularly, Microsoft added visual search to Bing and integrated ChatGPT into Skype. Hanging over it all is the specter of open-source AI software, which is rapidly approaching parity with the heavy hitters in the space. Earlier this year, an internal memo from a Google engineer said open-source software was their biggest threat in the AI market.
Open-source software would be more difficult for governments to regulate since projects are not limited to specific jurisdictions and can be contributed to by anyone with access to the internet. On the other hand, they also enable more transparency, with the source code freely available to the public.
There are a multitude of concerns that governments and members of the public have with AI. Cyber security experts fear voice cloning software and deepfake technologies could be used to circumvent biometric security measures. There are also concerns those same technologies could be used to spread misinformation.
Privacy experts have fears about AI-powered facial recognition software and how it can be used to gather information online. Others complain AI programs like chatbots and image-generation software were built on the intellectual property of millions of people.
But perhaps the most universal concern is economic. A recent report from Goldman Sachs said AI could potentially replace the work hours equivalent to 300 million full-time jobs, and the recent Writers Guild of America and the ongoing Screen Actors Guild strikes both listed AI-generated content as a concern.