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 has been unable to establish its AI credentials. Its own offering, the “experimental” chatbot Bard, compares poorly with rivals, and internal reports have portrayed a company in a panic and disarray. Today, at its annual I/O conference, the company needs to convince the public (and shareholders) that it has a meaningful answer. But to do that, you need a new playbook.
Google is undoubtedly a leader in AI investigation. As their executives like to point out, it was Googlers who created the transformative architecture that powers chatbots like ChatGPT. Equally significant, it was Googlers who called attention to the flaws in these systems (and, as a thank you, got fired). But Google has failed to do AI products; it has failed to take this work and mold it into tools that capture the public imagination. In short, the spirit of AI has been lost that, despite all the talk 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 sentiment arises from two main sources. The first is a technical ecosystem that is iterative and comparatively open. Several 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 quite closed, like OpenAI, push updates with impressive speed and offer attractive hooks for developers to develop with.
This leads to the second source: the products of these systems, which increasingly define the current cultural moment. if that is balenciaga harry potterthe boastful Pope, Deepfakes of President Joe Biden playing CS:GO, singers licensing AI voice clones to the publicor chatbots inspired by fan-favorite anime characters, there are thousands of cases of AI weirdness that entertain and sometimes infuriate.
It goes without saying that not all of these experiments are good. Many are malicious (like fake porn), and many more are simply irresponsible and poorly designed (like chatbot therapists). But the sum total of this work, good and bad, contributes to the sense of a shifting, protean technological ecosystem of change, experimentation, and cultural meaning. A tide that Google, despite all its experience, has completely missed.
This failure is best exemplified by Google’s work on AI language models and its Bard chatbot, especially when compared to the launch and track record of Microsoft rival Bing.
Today, talking to Bard is like being stuck in an AI nursery. If he strays too far from your acceptable question rate, he’ll receive a polite reprimand. “I’m sorry Dave. I’m afraid I can’t do that. Even when the system is helpful, the responses from it are insufferably bland. “Today, trees are an essential part of Earth’s ecosystems,” he told me in response to a question about the evolutionary history of trees. “They provide us with oxygen, food and shelter.” Sure, Bard. I guess. But also, why don’t you shoot me in the head while you’re at it?
Bing, by comparison, feels like the companion that helps you escape daycare That’s not to say he’s some semi-sentient entity or a perfectly crafted NPC. But the unpredictable edge of her responses creates the illusion of personality (grabbing hearts and headlines in the process), while her design encourages conversation rather than shuts it down.
This difference can be seen only in the basic UI options for the two chatbots. Bing, for example, consistently offers clickable feeds for its responses, which a) encourage exploration but also b) position the chatbot as more of a peer than an authority. It is open and permissive; it makes you feel like the system is somehow on your side as you navigate through the vast amount of information on the web.
Bard’s answers, by comparison, are much more independent. Occasionally the system offers links and citations, but the feeling is that Bard only offers access to its own domain, rather than serving as a portal to the Internet in general. It might not sound like much of a criticism, but the result is a muffled user experience, a conversation killer that has me climbing the featureless walls of Google’s bland Material You design. it’s just not fun.
This comparison is symptomatic of larger differences in Google’s and Microsoft’s approaches to AI. While Bard has been loitering (his refresh page shows only three changes since launch), Microsoft has been iterating rapidly, cramming chatbots into more and more of its products and rushing new features for Bing, from image generation to (coming soon) integration with apps like WolframAlpha and OpenTable. In short, you’ve been experimenting, and while your efforts may turn out to be misguided, at least you’re in tune with the moment.
I’m not sure what the answer is for Google here. I personally don’t think chatbots in their current form are a good replacement for search, full stop. As I’ve written before, issues like “hallucinations” are too persistent and damning to ignore. But in I/O, the company must show that it at least sees the potential: the excitement — of this technology. In the past, CEO Sundar Pichai has tried to talk, comparing AI to electricity or fire (nonsense, in my opinion), but such empty talk should be left to the bots. Instead, let’s see what humans can do.
