Mistral urges telcos to get into the hyperscaler game

Mistral urges telcos to build data centers, reduce US tech reliance.

: Mistral CEO Arthur Mensch called on telecom firms to invest in data centers at the Mobile World Congress in Barcelona. Emphasizing decentralizing the cloud and reducing European reliance on U.S. tech, he highlighted the partnership with Free for AI distribution. Mensch also discussed working with regulators on the EU AI Act and predicted more specialized AI models in the future.

Arthur Mensch, CEO of Mistral, addressed attendees at the Mobile World Congress, strongly urging telecommunications companies to invest in building their data center infrastructures to become hyperscalers. Hyperscalers are large cloud service providers that can scale up rapidly in response to user demand, and the market is currently dominated by U.S. giants like Amazon, Google, and Microsoft. Mensch suggested that a European investment in becoming hyperscalers would bolster the regional AI ecosystem, calling for more localized data centers to enable this shift.

Mensch emphasized the importance of decentralizing the cloud to make room for newer players in the field, suggesting that telcos and tech companies should work on reducing their reliance on U.S. technology. While acknowledging that some American technology infrastructure is inevitable, he advocated for strategic pragmatism. He also pointed out that fragmentation across the European Union's 27-member single market poses a bigger challenge than regulation for startups like Mistral.

To support his argument for telcos to engage with AI developments, Mensch highlighted that network operators must adapt their infrastructure to accommodate personalized data streams facilitated by AI. He hinted at potential collaborations for network upgrades and distribution partnerships in the AI consumer product realm, citing Mistral’s agreement with Free for the AI assistant Le Chat as an example.

Mensch also touched on the implications of AI regulation, considering the current state as workable but not ideal. He criticized the EU AI Act for being premature and overly focused on technology aspects while voicing optimism about changing perspectives among EU policymakers toward AI investment.

Looking ahead, Mensch predicted that AI models will become increasingly specialized, capturing human interactions to improve model accuracy and personalization. Mistral plans to leverage data from these interactions to develop robust and specialized AI systems designed for individual users over the next two years.

Sources: worldmobilecongress.com, TechCrunch, The Verge, Mistral.ai