Great language models aren’t new, but the instant popularity of OpenAI’s ChatGPT has pitted Microsoft against Google in the emerging Internet search war. This could even change the look of the network of networks.
Bing, Microsoft’s new chatbot powered by OpenAI technology as part of their multi-billion dollar partnership continues to function with apparent glitches. In fact, it has already become famous for its energetic and strange responses with users who have been able to try it.
But it looks like the hype is still high: Microsoft executive Yusuf Mehdi tweeted this week that “several million” people are on the waiting list to try the bot, which is already has reached 169 countries.
Meanwhile, Google appears to be working on pushing its own Bard AI chatbot, with CEO Sundar Pichai urging employees to work on getting the bot ready for deployment.
The creator of ChatGPT believes that “it won’t be that long before AIs are terrifying” and thinks that “regulation will be essential”
This technology promises to transform the market from what is now any mundane search that returns a page with a list of results. The tests that Business Insider What has been done with the new Bing, for example, have shown that it is an interactive process in which the bot can sometimes act as an assistant, quickly navigating the Internet to make specific travel and sightseeing recommendations or create daily agendas.
But that conversational model, AI assistant type, could be transferred to other activities that consumers do online like shopping or managing budgets or tracking physical exercise, reveals Prasanna Arikala, director of technology at the company Kore.ai, based in Orlando, USA.
The challenge, as OpenAI’s own leadership has recognized, will be to ensure that the robots provide reliable information.
“There is no doubt that it will change our way of life,” Arikala predicts to Business Insider. “But then, it’s very important to understand where you cross the Rubicon, the point of no return in terms of what to believe and how to consume that information responsibly.”
ChatGPT itself can also help train other AI modelsrecalls Cameron Turner, vice president of data science at Kin + Carta, a Chicago-based consultancy that advises companies on using data effectively to build their business. Turner also previously led a key data-driven team at Microsoft called the telemetry group.
Turner explains that ChatGPT could help produce organized data to transform information into something like feed that can feed other AI models into more specialized uses. AI technology is being developed, for example, to aid in healthcare research, fight cyber-attacks, and improve supply chain logistics.
Using AI in this way can help “if you want to experiment cheaply, without compromise, without making large-scale data purchases just to test a certain hypothesis,” he says.
ChatGPT is for idiots
And it’s not just about the Internet. Techies are also hopeful that OpenAI’s evolving language-processing tool could improve other types of cutting-edge technology that humans interact with. Could, for example, help robots learn to communicatesays Tesca Fitzgerald, an adjunct professor at Yale University who teaches artificial intelligence classes and researches robotics.
Robots are taught and programmed to move, sense, and interact with their environment, each of which requires specialized knowledge, Fitzgerald explains. OpenAI’s GPT technology can be used to help robots “talk” to the humans who run them, she says.
And he goes even further: “We could use the GPT to explain to robots how to break a problem into simpler steps to execute tasks.”