This week in AI was a week of reshuffling. SpaceX filed for its IPO at up to $2 trillion, Google unveiled its Gemini Spark personal agent, Meta cut 10% of its workforce, Microsoft canceled Claude Code licenses over runaway costs, and Andrej Karpathy announced he was joining Anthropic.
1. SpaceX files for IPO at up to $2 trillion valuation
SpaceX officially filed its IPO registration with the SEC on May 20, seeking to list on Nasdaq under the ticker SPCX. The company targets a valuation between $1.75 trillion and $2 trillion, with plans to raise between $40 billion and $80 billion. The filing reveals something significant: in 2025, SpaceX invested $12.7 billion in AI research and development, more than it spent on rockets or connectivity. In Q1 2026 alone, it spent $7.7 billion on AI. The company identifies a $28.5 trillion addressable market. The SpaceX IPO is not just a financial event; it is a signal that the company is repositioning itself as an AI company with rockets, not the other way around.
2. Google I/O 2026: Gemini Spark, the agent that works while your screen is off
At Google I/O 2026, Google unveiled Gemini Spark, a personal agent that operates 24/7 without needing an open computer. Spark integrates with Gmail, Docs, and the entire Workspace ecosystem, with third-party app support via MCP coming soon. Google also dropped the price of its Ultra plan from $250 to $100 per month and increased usage limits fivefold. The company announced Gemini 3.5 Flash and smart glasses developed with Samsung. With these moves, Google is positioning Gemini not as a chatbot but as a permanent automation layer in users’ lives.
3. Meta cuts 10% of its workforce and reassigns 7,000 employees to AI
Meta confirmed it is laying off 10% of its workforce, approximately 8,000 people, while reassigning another 7,000 employees to AI-focused teams. The company allocated $145 billion in capital expenditures for this year, its largest restructuring since 2022-2023. Mark Zuckerberg has been explicit: Meta will compete in AI at any cost. The message to the industry is clear: tech companies are redistributing human talent toward AI projects, and roles without a direct connection to that bet are at risk.
For sales teams tired of cold leads, slow customer responses, and manual processes, Dapta is the ultimate tool.
Dapta is the leading platform for creating AI sales agents specifically designed to increase inbound lead conversion. Respond to your leads in less than a minute with voice AI and WhatsApp that converts.
If you want your team to sell more while AI handles the complex stuff, you have to try it.
4. Microsoft cancels internal Claude Code licenses over runaway costs
Microsoft canceled Claude Code licenses in its Experiences and Devices division, effective June 30, 2026, just six months after launching the pilot program. Token-based billing consumed the annual AI budget within months. Uber reported a similar problem, warning it would exhaust its entire 2026 AI budget in four months. This case illustrates one of the main challenges of large-scale AI deployment: without usage governance, costs scale fast. Adopting AI agent tools requires predictable pricing models and per-team usage limits.
5. Andrej Karpathy joins Anthropic to lead pre-training
On May 19, Andrej Karpathy, OpenAI co-founder and former Tesla AI lead, announced he is joining Anthropic to lead pre-training work under Nick Joseph. His announcement post received 148,000 likes and 7,800 responses, reflecting the weight his name carries in the AI community. Karpathy is known for making AI accessible: his neural networks course has millions of views on YouTube. His arrival at Anthropic strengthens the technical team in the most strategic area of language model development and will likely accelerate the company’s fundamental research pace.
Bonus: Content of the Week
Replit grew from $2 million in annual revenue to $250 million, one of the fastest growth trajectories in startup history.
