Nvidia and Perplexity, an AI search company, have partnered with over a dozen AI firms across Europe and the Middle East.

Announced at an AI conference in Paris, the collaboration aims to enhance AI technologies and distribute them to local businesses.

The initiative will support model developers in countries like France, Germany, Italy, Poland, Spain, and Sweden.

The focus is on building reasoning models for complex tasks, especially in low-resource languages.

Nvidia will generate synthetic data and translate reasoning data for training. Perplexity will help distribute the trained models to European businesses.

.source-ref{font-size:0.85em;color:#666;display:block;margin-top:1em;}a.ask-tia-citation-link:hover{color:#11628d !important;background:#e9f6f5 !important;border-color:#11628d !important;text-decoration:none !important;}@media only screen and (min-width:768px){a.ask-tia-citation-link{font-size:11px !important;}}

🔗 Source: Reuters

🧠 Food for thought

1️⃣ Europe’s AI gap represents a substantial economic opportunity

The Nvidia-Perplexity partnership addresses a significant economic challenge for Europe, which has been lagging behind global AI leaders.

European firms currently capture only 12% of their digital potential, compared to 36% in the US, creating a substantial opportunity for growth through AI adoption 1.

McKinsey research indicates that effective AI development and diffusion could add €2.7 trillion to Europe’s economy by 2030, equivalent to 20% of its GDP, with an additional €900 billion possible if Europe catches up to the US AI frontier 1.

This economic potential explains why companies like Nvidia are investing in regional AI development, as successful AI deployment could transform productivity across multiple sectors, with retail alone potentially saving €300 billion in the coming years 1.

The partnership’s focus on local language models directly addresses one of Europe’s key barriers to AI adoption: the skills gap and need for accessible AI tools that can be implemented by businesses without specialized expertise.

2️⃣ Language-specific AI models are becoming strategic assets

Nvidia’s investment in developing local language models highlights how AI development is increasingly tied to linguistic and cultural needs.

The challenges of creating reasoning models in languages with less training data isn’t just a technical problem. It represents a potential competitive disadvantage for non-English speaking regions in the global AI race 2.

By generating synthetic data for “low-resource languages,” Nvidia is addressing a fundamental imbalance in AI development that has previously favored regions with larger volumes of digital text available for training 2.

This approach mirrors strategies in other regions, particularly the UAE and Saudi Arabia, where governments have established dedicated AI initiatives aimed at localizing technology development to serve their specific cultural and business contexts 3.

The economic impact of these language-specific models could be substantial, as research shows that three-quarters of firms using AI report increased revenues and productivity, suggesting that removing language barriers could accelerate adoption 4.

3️⃣ AI infrastructure is becoming a cornerstone of digital sovereignty

The partnership between Nvidia, Perplexity, and European AI firms reflects a growing global trend toward regional AI independence and data sovereignty.

Europe is actively building homegrown AI infrastructure to reduce its dependence on US providers, with NVIDIA supporting the deployment of over 3,000 exaflops of compute resources across France, Italy, Spain, and the UK 5.

The establishment of local AI ecosystems addresses both economic and regulatory concerns, as the EU AI Act—now in effect—sets global standards for ethical AI that require careful implementation by businesses operating in European markets 6.

For small and medium enterprises, which form the backbone of the European economy, localized AI models running in regional data centers could remove adoption barriers related to technical expertise and data privacy concerns that currently limit their AI utilization 7.

……