Microsoft calls for more companies to start using AI

Posted Nov 21, 2017 by James Walker
More companies have to deploy AI if it's to be broadly utilised in society, Microsoft has said. Although the tech is now receiving attention, the company warned it has to be applied to new fields if it's to avoid being controlled by a handful of firms.
Microsoft is launching new AI development tools to help businesses get started with the tech
Microsoft is launching new AI development tools to help businesses get started with the tech
Microsoft strategic policy advisor Dave Heiner discussed his perception of the risks of AI in an article with ZDNet. He explained that the technology must be used more broadly if it's to achieve its full potential. Limiting AI to a handful of industries could allow individual firms to dominate the technology.
Broad deployments
Microsoft is a strong supporter of AI adoption but has previously warned it mustn't be used to replace humans. Heiner said AI should "amplify human ingenuity," instead of directly supplanting it. A situation where a handful of companies are dominant in AI innovation could see this principle eroded. Unless AI gains a presence in every field, only a few people will stand to benefit from it.
"Any place [where] intelligence is helpful, which is just about every place, AI could be helpful as well," Heiner said during the Microsoft Summit in Sydney last week. "There's just no possibility whatsoever… if AI is just being run by four or five companies, that it can possibly be deployed broadly enough."
Nonetheless, there are several difficulties impeding the widespread use of AI. The computing resources and expertise required to build, train and utilise neural networks means it's hard for smaller firms to get established.
AI accessibility
AI research is still led by specialised institutes and technology giants such as Microsoft, Google and Facebook. With skilled AI programmers in short supply, startups with ideas for the future can find it hard to get their concept into development. This poses a big threat to the democratisation of machine learning.
Another challenge is in getting people to use AI-powered services. Most consumers are unfamiliar with how machine learning works or what happens to their data. Earning trust is an important consideration for companies that are experimenting with AI, particularly those using it to approach societal issues.
READ NEXT: Microsoft patents reveal conceptual mind control tech
It's not just consumers that lack understanding of AI. Companies using the technology might not necessarily appreciate how their machines have managed to solve a problem. Existing AI systems have already succeeded in learning new techniques on their own. Firms have to ensure they’ve fully audited the operation of their neural networks before they go live.
Even with these challenges, Microsoft called for more firms to start familiarising themselves with the tech, noting it will "augment our capabilities" and unlock new opportunities. The company said it wants to harness the hidden insights in the data reserves being generated globally, helping to improve society.