Google has had a rough six months. Since ChatGPT launched last November — followed by the new Bing in February and GPT-4 in March — the company hasn’t established its AI credentials. Its own offering, the “experimental” chatbot Bard, is doing poorly compared to rivals, and insider reports have depicted a company in panic and disarray. Today, at its annual I/O conference, the company must convince the public (and shareholders) that it has a meaningful answer. But to do that, it needs a new roadmap.
Google is undoubtedly a leader in AI research. As the executives like to point out, it was Googlers who created the transformer architecture that powers chatbots like ChatGPT. Just as importantly, it was Googlers who drew attention to the shortcomings of these systems (and were fired in thanks). But Google has failed to create AI Products; it has failed to take this labor and transform it into tools that capture the imagination of the public. In short, it lacks the AI zeitgeist, which – for all discussions of existential risk and economic threat – is Also defined by a sense of exploration, experimentation and creative, chaotic fun.
AI art and tools are increasingly defining the current cultural moment
This feeling stems from two main sources. The first is a technical ecosystem that is iterative and relatively open. A number of important AI models are open source (such as Stable Diffusion); many more are shared or leaked (such as Meta’s LLaMA language model). Even companies that are pretty closed-minded like OpenAI push updates at an impressive rate and provide enticing hooks for developers to build on.
This leads to the second source: the outputs of these systems, which increasingly determine the current cultural moment. Whether that’s Balenciaga Harry Potter, the stripped-down pope, deepfakes of President Joe Biden that CS:GO, singers licensing AI voice clones to the public, or chatbots modeled after fan-favorite anime characters, there are thousands of instances of AI weirdness that entertain and sometimes infuriate.
It goes without saying that not all of these experiments are Good. Many are malicious (such as deepfake pornography), and many more are simply irresponsible and poorly designed (such as chatbot therapists). But the sum of this work – good and bad – adds to the sense of a swirling, multi-faceted technological ecosystem of change, experimentation and cultural significance. A tide that Google, for all its expertise, has completely missed.
This failure is best illustrated by Google’s work on AI language models and its chatbot Bard, especially when compared to the launch and trajectory of Microsoft’s rival Bing.
These days, talking to Bard feels like being stuck in an AI nursery. Stray too far from the index of acceptable questions and you will be politely reprimanded. ‘I’m sorry, Dave. I’m afraid I can’t.” Even if the system is useful, the answers are insufferably bland. “Today, trees are an essential part of Earth’s ecosystems,” it told me in response to a question about the evolutionary history of trees. “They provide us with oxygen, food and shelter.” Sure, Bard. I think so. But why don’t you shoot me in the head too while you’re at it?
Bing, on the other hand, feels like the sidekick helping you out escape childcare. That’s not to say it’s a semi-sentient entity or seamlessly crafted NPC. But the unpredictable side of the answers creates the illusion of personality (winning hearts and headlines), while the design encourages conversation rather than closing it down.
This difference can only be seen in the default UI choices for the two chatbots. For instance, Bing consistently provides clickable resources in its answers, which a) encourage exploration, but also b) position the chatbot as something closer to a companion than an authority. It is open and tolerant; it makes you feel like the system is somehow on your side as you navigate the vast amount of information on the web.
Bard’s answers, on the other hand, are much more self-contained. The system occasionally provides links and quotes, but the sense is that Bard only provides access to its own domain, rather than functioning as a portal to the wider internet. It may not sound like a big criticism, but the result is a numb user experience; a conversation killer that has me crawling past the bland walls of Google’s smooth Material You design. It’s just not pleasure.
This comparison is symptomatic of greater differences in Google’s and Microsoft’s approaches to AI. While Bard has been lazing around (the update page only shows three changes since launch), Microsoft has been quickly iterating, putting chatbots into more and more of its products and launching new features for Bing, from image generation to (coming soon) ) integration with apps like WolframAlpha and OpenTable. In short, it’s been experimenting, and while its efforts may prove misguided, it’s at least tuned for the moment.
I’m not sure what the answer for Google is here. Personally, I don’t think chatbots in their current form are a good replacement for search, period. As I’ve written before, things like “hallucinations” are just too persistent and devastating to ignore. But with I/O, the company has to prove that it at least sees the potential tension – of this technology. In the past, CEO Sundar Pichai has tried to make the conversation by likening AI to electricity or fire (which is a silly thing to say, in my opinion), but such empty talk should be left to the bots. Instead, let’s see what the people can actually make.