Siri and Google Assistant Turn to AI for Improvements

Next week’s WWDC 2024 keynote carries significant weight. The stakes for this presentation surpass typical post-event market reactions. Tim Cook and his team are under immense pressure to deliver, even more so than before last year’s Vision Pro announcement.

On Monday, Apple will reveal its AI strategy. This topic has been a major point of speculation in Cupertino over the past few years, especially as competitors like Google and Microsoft have adopted generative AI. The industry widely agrees that large language models such as ChatGPT and Gemini will drastically change how we interact with our devices.

Apple is anticipated to announce a partnership with OpenAI, integrating its technology into iPhones and Macs. The company’s immediate plan focuses on deeply embedding generative AI within its existing ecosystem, centering around Siri. Since Siri’s launch in 2011, Apple has aimed to make the voice assistant a core feature across all its operating systems.

However, in the 13 years since its debut, Siri has not fulfilled the revolutionary promises Apple made. There are many reasons for this, with the primary being its capabilities. The idea of an artificial voice assistant existed long before Siri, but no one perfected it due to the complex demands placed on these assistants as smartphones evolved.

Despite the remarkable efforts by the Stanford Research Institute, the technology for a seamless experience wasn’t available. Siri co-founder Norman Winarsky acknowledged this in 2018, explaining that Apple initially aimed for a more limited assistant focusing on areas like entertainment and travel. “These are hard problems, and when you’re a company dealing with up to a billion people, the problems get harder yet,” Winarsky said. “They’re probably looking for a level of perfection they can’t get.”

Generative AI hasn’t reached that level of perfection yet either. Issues like hallucinations still persist. Despite the excitement of recent years, it feels like we’re still in the early stages. Google, for instance, has been aggressive in some areas, the best example being the company’s decision to feature Gemini results prominently in search queries.

When a dominant search engine prioritizes unproven resources, it must aim for near-perfect accuracy, rather than giving users problematic advice like eating glue. Google tags Gemini results as a product of its “Search Labs,” but many users don’t grasp the maturity level this implies or bother to seek more information.

I’ve spoken with numerous researchers over the years who describe the results from large language models as “magic,” highlighting our limited understanding of this technology despite the incredible advancements happening in the space.

As Arthur C. Clarke famously said: “Any sufficiently advanced technology is indistinguishable from magic.”

Google has taken a more measured approach by integrating Gemini into Android instead of replacing Assistant outright. Users can choose to make Gemini their default by assigning it to the Assistant button on Pixel devices, but this requires a deliberate action so far.

Although Gemini has not yet fully overtaken Android, Google suggests a future where it might replace the Assistant entirely. An announcement along those lines was expected at I/O last month, but Google wisely chose to let Gemini develop further.

The decision to retain the Assistant name is ultimately about branding. Apple, similarly, is committed to the Siri brand, having marketed it for over a decade. Nonetheless, generative AI will soon dominate the smart assistant space.

Voice assistants overall are facing an existential crisis. The demand for smart speakers, a barometer for platforms like Siri, Alexa, and Google Assistant, has declined after a pandemic boom. While it’s premature to declare the category doomed, it needs significant innovation to survive in the long term.

Generative AI seems positioned to be the next big thing, but early hardware like the Humane Ai Pin and Rabbit R1 shows how far there is to go before it’s a mainstream experience.

On Monday, Apple will finally reveal its strategy. With rumors suggesting the company shifted many employees to focus on generative AI after its electric car project faltered, Apple appears to have allowed competitors a head start. Partnering with a leader like OpenAI seems the most logical step.

When Siri’s acquisition was announced, Steve Jobs was asked if Apple aimed to outdo Google. “It’s an AI company,” Jobs replied. “We’re not going into the search business. We don’t care about it. Other people do it well.”

Apple’s current strategy for generative AI reflects a similar stance. Rather than trying to outdo OpenAI, Apple is collaborating. Yet, even the top models need more development before they can replace today’s smart assistants.