Microsoft: ‘ever present’ AI assistants are coming

Microsoft’s head of AI, Mustafa Suleyman, recently shared insights into the future of artificial intelligence, predicting that advanced AI assistants with “highly effective long-term memory” could become a reality within a year. These products aim to recall conversations, projects, and problems, enhancing user engagement and encouraging individuals to share more of their personal history with these digital companions.

Suleyman envisions a shift towards a new era where these persistent and capable AI co-pilots become integral to daily life. However, this level of integration raises significant concerns. Critics worry about data security, privacy, and the potential for AI tools to provide inaccurate advice or exhibit bias against the users they are designed to assist. Proponents argue that for AI to be genuinely helpful, it must be deeply embedded in our lives, possessing context and historical knowledge to fulfill tasks effectively.

For instance, an AI diary manager would need access to your calendar, enabling it to organize your schedule based on your activities. Suleyman contends that people’s privacy expectations are evolving. He points to the prevalence of devices—such as TVs, laptops, and smartphones—that are “recording continuously” in everyday environments. He also cited features like Apple’s Live View, which records audio and video simultaneously when taking photos, suggesting that many users embrace this technology. “Most people love that feature,” he noted, emphasizing that perceptions of privacy are changing.

The acceptance of such technology hinges on users’ ability to control its use and their trust in the providers. Suleyman highlighted that the benefits of these advancements will significantly influence whether people embrace them.

A Paradigm Shift or Passing Trend?

Microsoft’s substantial investment in OpenAI, the creator of ChatGPT, has positioned the company as a frontrunner in the race to harness AI technology. However, recent research indicates that widespread adoption may not be as straightforward. A poll by the Reuters Institute revealed that while 29% of respondents in the UK had used ChatGPT, only 2% used it daily.

Suleyman acknowledged the possibility that consumer AI tools might not achieve the same universal popularity as smartphones. “Maybe this is different from the smartphone,” he remarked, speculating that while nearly 90% of the global population uses smartphones, AI tools may face rejection from around 50% of users.

Despite these uncertainties, Suleyman pointed out that AI has been the fastest-growing technology in history, regardless of its risks. He firmly dismissed the notion, proposed by industry analysts like Jim Covello from Goldman Sachs, that AI might turn out to be a passing fad similar to previous tech bubbles. Suleyman shared a story about an entrepreneur who successfully launched her business with support from an AI chatbot he developed, illustrating the practical, transformative applications of AI.

“We are clearly producing personalized, interactive knowledge at your fingertips at zero marginal cost,” he asserted. “The idea that this could be a bubble is utterly beyond me.”

New Developments from Microsoft

In line with its ambitious vision, Microsoft recently introduced a suite of new features for its AI assistant, CoPilot. These enhancements include a voice function, a daily news digest, and a specialized chatbot called “Think Deeper,” designed for complex queries. Another notable addition is CoPilot Vision, which will integrate into the Edge web browser, providing users with assistance while browsing. This tool is designed to enhance online activity without recording or storing data, requiring manual activation and automatically closing at the end of each session.

Microsoft has opted to limit CoPilot Vision to specific websites, with no official release date announced yet. Additionally, the company previously paused the rollout of another AI tool called Recall, which captured screenshots to help users retrieve information, following concerns from privacy advocates and inquiries from the UK’s data protection authority.

As Microsoft forges ahead with its AI initiatives, the balance between innovation and user privacy remains a pivotal discussion point. The evolution of AI assistants may soon redefine how we interact with technology, but careful consideration of ethical implications will be essential as these tools become more integrated into our lives.