Luis Poveda's AI Newsletter: Jun 16, 2025

In partnership with

Learn AI in 5 minutes a day

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

1. Sign up for The Rundown AI newsletter

2. They send you 5-minute email updates on the latest AI news and how to use it

3. You learn how to become 2x more productive by leveraging AI

Sign up to start learning.

This week witnessed a fascinating contrast in AI development strategies, with European AI powerhouse Mistral launching its first reasoning model Magistral under an open-source license, while Silicon Valley giants pursue dramatically different paths. Meta made headlines with a massive $14.3 billion investment in Scale AI and Mark Zuckerberg's formation of an ambitious "superintelligence" team offering compensation packages up to $100 million to top researchers. Meanwhile, Apple faced disappointment at WWDC 2025 as the promised AI-powered Siri upgrade remained notably absent, highlighting the company's cautious approach amid fierce competition from rivals who continue to push aggressive AI breakthroughs.

Mistral AI Launches Magistral: Europe's Answer to OpenAI's Reasoning Models

European AI leader Mistral AI entered the competitive reasoning model arena with the launch of Magistral, marking a strategic dual release that reinforces the company's commitment to open-source development. The announcement features two variants: Magistral Small, a 24-billion parameter model released under the permissive Apache 2.0 license, and Magistral Medium, a more powerful enterprise-focused version. This move directly challenges OpenAI's o3 series and DeepSeek's reasoning models while maintaining Mistral's foundational open-source philosophy. Magistral demonstrates impressive performance on mathematics benchmarks, scoring 73.6% on AIME-24 and jumping to 90% accuracy with majority voting techniques. The model's transparent "chain-of-thought" reasoning and multilingual capabilities in French, Spanish, German, Italian, Arabic, Russian, and Chinese position it as a compelling alternative for enterprises requiring verifiable AI decision-making processes.

• VentureBeat: Mistral's first reasoning model, Magistral, launches with large and small Apache 2.0 version

• Axios: Mistral launches new AI reasoning models to compete with OpenAI

Meta's $14.3 Billion Scale AI Bet Signals Desperate Catch-Up Strategy

Meta executed one of the largest AI investments in recent history with its $14.3 billion acquisition of a 49% stake in data-labeling company Scale AI, valuing the startup at $29 billion. The deal includes the recruitment of Scale AI's co-founder and CEO Alexandr Wang to join Meta's newly formed "superintelligence" team, as part of Mark Zuckerberg's ambitious effort to close the AI gap with rivals Google, OpenAI, and Anthropic. This strategic move addresses Meta's urgent need for high-quality training data as the company reportedly lost 4.3% of its top AI talent to competing labs last year. Scale AI, which has transitioned from serving autonomous vehicle companies to becoming a critical infrastructure provider for major AI labs, expects $2 billion in revenue this year primarily from data labeling services that power large language models.

• TechCrunch: Scale AI confirms 'significant' investment from Meta, says CEO Alexandr Wang is leaving

• Forbes: Meta Could Spend Majority Of AI Budget On Scale AI As Part Of $14 Billion Deal

Zuckerberg Assembles $100 Million "Superintelligence" Dream Team

Mark Zuckerberg is personally leading the formation of an elite AI research team at Meta focused on developing artificial intelligence that surpasses human capabilities, offering compensation packages as high as $100 million to attract top researchers from competing labs. The ambitious "superintelligence" initiative represents Meta's most aggressive play yet in the AI arms race, targeting the recruitment of approximately 50 world-class researchers to build what the company describes as "smarter-than-human artificial intelligence." This effort goes beyond the current industry focus on artificial general intelligence (AGI), which aims to match human cognitive abilities, instead pursuing the more ambitious goal of creating machines more powerful than the human brain. The initiative reflects Zuckerberg's determination to recover from recent setbacks and defections that have left Meta trailing competitors in the race for AI dominance.

• Morning Brew: Zuck is assembling an AI 'superintelligence' team

• The New York Times: Meta Is Building a Superintelligence Lab. What Is That?

Apple's WWDC 2025 AI Disappointment Exposes Strategic Vulnerabilities

Apple's much-anticipated Worldwide Developers Conference failed to deliver the promised AI-powered Siri upgrade, leaving investors and developers disappointed as the company announced only incremental Apple Intelligence improvements. The tech giant's cautious approach contrasted sharply with competitors' aggressive AI strategies, focusing on practical features like live phone call translation and visual search rather than breakthrough AI capabilities. Apple shares closed down 1.2% following the announcement, reflecting investor concerns about the company's ability to compete in the rapidly evolving AI landscape. The absence of the more personalized Siri, first promised at last year's WWDC, highlights ongoing challenges within Apple's AI division that led to an executive reshuffle earlier this year. While Apple did announce opening access to its on-device AI language model for third-party developers and integrating ChatGPT into Xcode, industry analysts characterized the overall AI announcements as "incremental at best."

• Ars Technica: Apple tiptoes with modest AI updates while rivals race ahead

• TechRadar: AI Siri's absence from WWDC shows it's still Apple's biggest failure

Zuckerberg Doubles Down on AI with Billions on the Table

Mark Zuckerberg isn’t playing around. With a massive $14.3 billion acqui-hire of Scale AI, and reports of compensation packages reaching up to $100 million for top-tier AI researchers, it's clear: Zuck is putting his money where his models are.

Meta’s LLMs have shown impressive results in recent benchmarks, but they’re still struggling to claim the top spot on the leaderboard. That might be all the motivation Zuckerberg needs to push harder, and faster, into the AI arms race.

Meanwhile, over in Cupertino, Apple continues to fumble the AI ball. The Siri upgrade they hyped last year? Still missing in action. If things don’t turn around soon, the calls for Tim Cook’s resignation may start getting louder and harder to ignore.

This week's developments reveal a defining moment in the AI industry, where traditional Silicon Valley dominance faces challenges from multiple fronts. Mistral's open-source Magistral demonstrates that European AI companies can compete with American giants while maintaining a commitment to transparency and accessibility. Meanwhile, Meta's desperate billion-dollar investments and Zuckerberg's superintelligence ambitions highlight the intense pressure facing established tech companies to maintain relevance in the AI race. Apple's conservative approach and technical stumbles at WWDC suggest that even the world's most valuable company isn't immune to AI disruption. As the industry moves beyond current language models toward reasoning systems and superintelligent machines, the next phase of competition will likely be determined by which companies can most effectively combine massive capital investments, top talent acquisition, and breakthrough technical innovations.

Start learning AI in 2025

Keeping up with AI is hard – we get it!

That’s why over 1M professionals read Superhuman AI to stay ahead.

• Get daily AI news, tools, and tutorials

• Learn new AI skills you can use at work in 3 mins a day

• Become 10X more productive

Sign up and start mastering AI