‘AI is not simply..’: Anand Mahindra backs Palantir CEO’s idea, sees India’s edge lie elsewhere
As nations and corporations pour billions into constructing ever extra highly effective synthetic intelligence fashions, Anand Mahindra believes the {industry}’s greatest long-term benefit could lie elsewhere.
Reacting to a current interview by Palantir CEO Alex Karp, the Mahindra Group chairman argued that whereas frontier AI fashions stay essential, the true industrial and strategic edge will more and more come from the appliance layer — the software program constructed on prime of these fashions that makes AI usable, safe and useful for companies.
What Anand Mahindra stated
In a publish on X (formally twitter), Mahindra stated Karp’s interview would “be broadly mentioned and debated,” acknowledging that some could view the Palantir chief’s feedback as self-serving. Nevertheless, he stated the underlying framework deserves critical consideration.
In line with Mahindra, AI consists of three interconnected layers:
- Compute — the computing infrastructure powering AI.
- Fashions — the big language fashions that generate intelligence.
- Purposes — software program that turns these fashions into sensible enterprise options.
Mahindra agreed with Karp’s argument that essential infrastructure can’t depend on AI fashions alone.
“Crucial infrastructure does not run on a mannequin alone; it wants an utility layer sitting on prime,” he wrote.
‘AI is sort of a smartphone’
Drawing from his current shareholder letter to Tech Mahindra, Mahindra in contrast in the present day’s AI panorama to the smartphone revolution. “A smartphone is exceptional know-how,” he famous, “however it turned indispensable due to the apps and experiences constructed on prime of it.”
Equally, he argued, lasting worth in AI won’t essentially be created by the businesses constructing the underlying fashions or chips, however by these growing specialised functions that resolve real-world enterprise issues.
“The ecosystem determines who creates lasting worth, not the chip or the mannequin beneath,” he stated.
Why the appliance layer issues
Mahindra highlighted Karp’s argument that enterprise AI requires safeguards past what a basis mannequin can present.
In line with Karp, the appliance layer makes AI “secure and exact” by addressing questions that companies and governments can’t ignore, together with:
- Who owns the information?
- The place is the information saved?
- Are prompts safe?
- Is info being transferred externally?
- Can choices be audited and ruled?
These capabilities grow to be particularly essential in sectors comparable to defence, categorised authorities programmes, banking, healthcare and different extremely regulated industries.
India ought to nonetheless construct frontier AI fashions
Regardless of emphasising functions, Mahindra clarified that he continues to assist India’s efforts to develop sovereign frontier AI fashions.
Nevertheless, he instructed that if AI fashions more and more grow to be commoditised — with a number of open and proprietary fashions out there — the nation’s technique mustn’t concentrate on fashions alone.
As an alternative, he argued that compute, fashions and functions ought to evolve collectively, permitting India to construct a whole AI ecosystem moderately than counting on any single layer.
Why AI service firms may gain advantage
Mahindra believes AI service corporations could get pleasure from a sturdy aggressive benefit as a result of they already possess many years of enterprise workflow experience.
Somewhat than proudly owning the underlying mannequin, these firms can stay model-agnostic — selecting whichever open or proprietary mannequin most closely fits a buyer’s wants — whereas constructing safe, industry-specific functions that protect enterprise information, governance and operational management.
“The appliance additionally has to allow enterprise enterprises to protect their ‘alpha’,” Mahindra wrote, including that that is the place AI service firms are greatest positioned to create long-term worth.
