• Starter AI
  • Posts
  • Mistral's $640M raise, AI’s generational learning, Google’s environmental challenges

Mistral's $640M raise, AI’s generational learning, Google’s environmental challenges

Europe's biggest gets a boost.

Hello, Starters!

A lot is happening in the AI world, and with the sudden wave of news, it's easy to get lost. As usual, we've gathered the most relevant insights. Have a look!

Here’s what you’ll find today:

  • Mistral secures impressive $640M funding

  • Generational learning for AI agents

  • Does Google's AI ambitions harm the environment?

  • Adobe faces major backlash

  • Elon Musk drops his OpenAI lawsuit

  • And more.

Mistral is one of the most promising AI startups at the moment, as well as the main leader in Europe's AI landscape. If that statement isn't enough, let the numbers speak: they have secured $640 million in Series B funding, skyrocketing their valuation to nearly $6 billion.

Big names such as IBM, Samsung, Salesforce, Andreessen Horowitz, Cisco, and Nvidia are among its investors. Mistral plans to use this financial push to increase its compute capacity, aiming to become a global force and take OpenAI's current spot.

Lessons are generational, even for AI agents. Research from Google DeepMind and the University of Oxford has disclosed how cultural accumulation—a main building block in our knowledge and skills—can be achieved by agents through reinforcement learning, outperforming solo learners.

They use two RL models: an "In-Context" model, which accumulates knowledge by learning from other agents and improves itself to provide useful demonstrations to the next generation, and an "In-Weights" model, where accumulation happens more slowly as it is based on full training cycles, with models trained from scratch but benefiting from previous observations.

As Google continues to work on its "AI Overviews," an important question remains: are users considering the financial and environmental costs of AI search? If so, what's Google's approach to overcoming this potential problem?

Despite its benefits, AI technology introduces new challenges when discussing environmental issues, as a simple query to a model consumes as much energy as a regular smartphone charge. With most big tech companies continuing to build data centres and expand their presence in AI, many specialists are worried that their efforts to find sustainable options, such as renewable energy, aren't enough.

😨Adobe recently faced significant backlash from its users after updating its terms and conditions with an ambiguous phrase that led many to believe they were using private material to train its generative AI. The company has now issued a statement addressing the concerns, as most customers threatened to cancel their subscriptions.

🧑‍⚖️In a sudden twist of circumstances, Elon Musk has dismissed the lawsuit he initially filed in February against OpenAI, claiming that the startup had abandoned its original mission of developing AI for the benefit of humanity rather than for profit. There are no clear reasons why the CEO and his attorneys made this move.

What did you think of today's newsletter?

Login or Subscribe to participate in polls.

Thank you for reading!