Microsoft's head of AI, Mustafa Suleyman, has warned that unchecked development of artificial intelligence could give rise to a new "silicon species" capable of rivalling humans, unless the industry puts stronger safeguards in place.
Speaking to the BBC's Today programme on Thursday, Suleyman said rival firm Anthropic was taking a misguided approach by treating its AI as though it were human, warning that this could produce technology humanity would be unable to control.
He said it was reasonable for people to be concerned about AI, but added that there were practical steps that could be taken to keep it in check. Anthropic did not immediately respond to the BBC's request for comment.
Suleyman cautioned against AI firms building systems capable of setting their own objectives, earning money and owning assets, arguing this risked creating machines that would compete with humans for resources regardless of how much they claimed to value humanity.
His remarks followed an essay published earlier in the week in which he questioned Anthropic's approach to training its Claude model, even as he praised chief executive Dario Amodei and his team as intellectually honest.
In the essay, Suleyman argued that encouraging human-like qualities in Claude, a practice known as anthropomorphising, risked making the system appear to have its own desires and sense of self, when in fact, he wrote, AI systems do not feel, experience or suffer, and have no innate preferences.
ALSO READ: Threat Actors Tried Using Claude AI To Make Bioweapons, Anthropic Reveals In New Report
The Microsoft executive called for greater transparency in how AI systems are trained and evaluated, along with independent scrutiny of AI behaviour and stronger tools to monitor and control the technology.
Dame Wendy Hall, professor of computer science at the University of Southampton, told the BBC that Suleyman's comments were the sort of conversation the world needed to be having internationally, contrasting them with what she called the histrionics coming from some AI companies.
Suleyman said AI development needed to prioritise alignment, so that advanced systems remain subordinate to humanity, adding:
"Everybody must be aligned." Microsoft, which set up its own superintelligence team in October last year, published a draft Humanist AI Code of Conduct on Monday, setting out its vision of "humanist superintelligence" — AI it describes as highly advanced but one that stays within limits and remains under human control.
ALSO READ: Novo Nordisk Taps Anthropic's 'Claude Science' To Accelerate Drug Discovery
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.