Artificial intelligence might be about to change the way the internet works forever, according to its proponents.
The mainstream popularity of conversational AI system ChatGPT, and its integration into Microsoft’s Bing, are a peek at the kinds of changes that supporters of artificial intelligence have long suggested are coming.
For now, the systems are focused primarily on helping people to find information, by quickly searching through a vast corpus of text and creating an answer. Both ChatGPT and the new Bing are about gathering information from the internet and presenting it to users in a sometimes new, sometimes old form.
For years, however, the job of sorting through the internet’s information has been the domain of Google. “Our mission is to organise the world’s information and make it universally accessible and useful,” Google’s ‘about’ website reads.
It has repeatedly indicated that it sees the future of that mission as being in AI, and has said that it is working on products to do so. But, very quickly, it appears to be getting overtaken by other companies: not just OpenAI and its partners at Microsoft, but other companies too.
Google announced in 2017 during its annual developer conference that it would be moving from a “mobile first world to an AI first world”. Its chief executive Sundar Pichai, then new to the job, described it as an “important shift” that would guide how the company worked from that point on.
That year Google showed off features that were informed by its work on artificial intelligence. In 2017, Mr Pichai pointed to Google Lens, for instance, which is able to recognise places or objects in an image and tell users more about them.
“It’s inspiring to see how AI is starting to bear fruit that people can actually taste,” he wrote in a 2017 blog post summarising those new features. “There is still a long way to go before we are truly an AI-first world, but the more we can work to democratise access to the technology—both in terms of the tools people can use and the way we apply it—the sooner everyone will benefit.”
In the time since, Google has shown off yet more of that fruit, with new features and updates informed by its work on AI. It has also been working privately on what seemed to be more advanced AI that it opted not to publicly release, such as the LaMDA system that is mostly famous for having convinced one of Google’s own engineers that it had become sentient.
And some of that work has helped build the foundation for the other projects away from Google that regularly go viral – as Mr Pichai was quick to remind the world in his blog post announcing the development of Bard.
“Advanced generative AI and large language models are capturing the imaginations of people around the world. In fact, our Transformer research project and our field-defining paper in 2017, as well as our important advances in diffusion models, are now the basis of many of the generative AI applications you’re starting to see today,” he wrote.
But while Google might have helped build those products, it is yet to actually launch one of its own with the success of its competitors. It has had AI-powered experiences go viral in the past – such as its psychedelic ‘Deep Dream’ and a tool that compared people with famous paintings – but the new wave has been conspicuously free of Google’s input.
In Google’s telling, that is at least in part because it is being so careful about the dangers of those AI products. Internally, the company is reported to have suggested that it is behind other companies on AI because it is looking to be responsible, ensuring that its products are safe to release before they are made public.
This week, however, it announced “Bard”, a new system that looked specifically to be an attempt to take on the challenge from ChatGPT and other systems. The tool will be a new way of doing search, it suggested, offering information and inspiration to people.
Even in its announcement of Bard, it looked to stress the kinds of limitations and safety features it will be using, once again echoing its argument for why it appeared to be lagging behind. “We’ll combine external feedback with our own internal testing to make sure Bard’s responses meet a high bar for quality, safety and groundedness in real-world information,” Google wrote in its blog post.
It does not seem to be enough. In that same announcement, Google showed an example answer from Bard that was actually wrong, making a mistake about the James Webb Space Telescope.
The company also held an event that once again attempted to show its work on AI, and was once again met with a mostly tepid response. Taken together, the scepticism sent its shares plunging, and wiped $100 billion from Google’s valuation.
At the same time, Microsoft’s shares rose. And social media was filled with jokes about how the company had done the possible: made Bing a search engine that people were actually excited to use.
“Imagining a Bing exec somewhere deep in a Microsoft office just sobbing uncontrollably [right now] that Bing is finally getting its moment,” wrote Engadget journalist Karissa Bell.
