Microsoft hires DeepMind co-founder amid Google-Apple news

Microsoft's needs to be the AI company

Microsoft's news comes as the vendor strives to outdo its chief AI competitor, Google.

Google acquired DeepMind in 2014. Since then, the startup has become known as one of the leading AI research organizations.

DeepMind was a precursor to what is now another of Google's main rivals, OpenAI, the Microsoft partner and creator of the ChatGPT, GPT-4 and Sora generative AI models.

OpenAI's CEO Sam Altman was previously offered the new AI position at Microsoft after internal struggles with OpenAI's board members.

"Microsoft has been trying to get skilled AI leaders for a while," said Jack Gold, an analyst at J. Gold Associates.

This goal shows how the tech giant is continually trying to prove it has better AI technology than Google, Gold said.

"Microsoft has got a real issue in trying to convince the market that their AI systems are going to be better than anybody else's," he said. "Right now, we don't know. It's too early."

There are similarities between what Microsoft is now doing in the AI market and what it attempted with its Bing search engine, Gold continued. Although Microsoft invested millions in Bing, it could not beat Google's search engine, which dominates the search market.

"The same thing is going to happen in AI if Copilot doesn't do what people expect it to do," Gold said.

However, the move to hire Suleyman and Simonyan shows that the key to winning in the AI race is as much about having the right talent on the team as it is about having the right technology, Futurum Research analyst Ron Westfall said.

"Microsoft is making a smart move because it's also about acquiring the skills or the talent that's needed to help differentiate, 'Why Copilot? Why to go with Microsoft and OpenAI versus Google Gemini?" Westfall said, referring to Google's main generative AI system.

It also shows that the previous controversy with Altman is resolved and Microsoft is now looking ahead to how it can stay above its competitors, Westfall added.

Google and Apple

However, Microsoft is not the only cloud provider to display a competitive streak this week. Google is in talks to incorporate its chatbot and LLM Gemini into Apple iPhones, according to Bloomberg.

While neither Apple nor Google have confirmed it, the prospect proved to be a winning mark for Google and Apple in the stock market. Google's shares were up more than 4% on Monday. Apple's shares also saw a jump.

"It's a very promising combination," Opus Research analyst Dan Miller said. "[Apple chatbot] Siri needs a generational boost. Google needs to expand its presence on iPhones beyond search."

Microsoft has got a real issue in trying to convince the market that their AI systems are going to be better than anybody else's.

For the past few years, Google has been the default search engine on all Apple iOS devices. However, that arrangement is currently being examined by the Department of Justice.

With the addition of Gemini, the existing partnership between the two tech giants would expand, Miller added.

Adding Gemini to its iPhone will give Apple yet another competitive edge in the smartphone market that it already leads, Westfall said.

Some think the iPhone, while the most popular smartphone with consumers, falls behind some Android phones, such as Google's Pixel 8 or Samsung Galaxy, in features and technology. Apple is also seen as lagging in the early stages of on-device generative AI capabilities, Westfall added.

"What this indication here is that Apple cannot wait any longer to get on board with a headline grabbing generative AI deal for the iPhone," he said.

However, Apple is not likely to rely solely on Google. The company has its own generative AI projects, such as its large language model Ajax and Apple GPT.

If a deal between Apple and Google is actualized, Apple could use Gemini for on-cloud training and its own on-device generative AI capabilities, Westfall continued.

"The licensing arrangement could end up being a bridge for Apple," he said. He added that while onboarding the Gemini capabilities, Apple might spend more time and portfolio resources to build its own proprietary LLM.

Google Gemini challenges

For Google, the possibility of an expanded partnership with Apple has served to defuse some of the problems it's facing with the Gemini LLM. The cloud provider recently was forced to turn off the image-generating feature of Gemini after it produced historically inaccurate features.

It also recently revealed that the Gemini chatbot will not answer any questions about the U.S. elections.

However, Google's restriction of the model does not solve the original problem of bias in its flagship LLM, said Gary McGraw, cofounder of the Berryville Institute of Machine Learning.

"The issue is that they trained up the foundation model on the polluted ocean and now they're trying to stop the pollution from getting out with a filter, and that doesn't work," he said. "These models were built by drinking a data ocean without cleaning it first. And we have to do better than that." And Microsoft has the same problem, he added.

What this means for Apple is that if it uses Gemini on the iPhone and the model does something wrong, the iPhone maker could divert blame, Gold said.

"Their argument would be, 'Look, we're just a device. We are just the platform,'" he said. "It's going to be an interesting environment, because God forbid AI does something that causes harm to somebody. That's when people really get sued."

Apple is also likely to insist on not sharing data, Forrester Research analyst Thomas Husson said.

However, whether it be Apple, Microsoft or Google, the generative AI market is still new, Gold said.

"We don't know where the market is going to go," he said. "The market for Microsoft and Google and everyone else is, 'How do I broaden my AI stuff enough so that I can cover all this terrain going forward?'"

Esther Ajao is a TechTarget Editorial news writer and podcast host covering artificial intelligence software and systems.