Zuck's personally recruiting a team to chase down ‘superintelligence’
June 11, 2025
Good morning. Timbaland just launched an AI music label with a virtual artist named TaTa—weeks after hosting an independent artist competition that fans suspect was just "talent farming" to train his algorithms. His defense? A bizarre video of AI animals telling critics to stop crying because "AutoTune is AI too."
— The Deep View Crew
In today’s newsletter:
🏥 AI for Good: UH rolls out AI for precision cancer care
🎇 Trump administration preps AI.gov platform for July 4 launch
🧠 Zuck’s building a team to chase down ‘superintelligence’
🏥 AI for Good: UH rolls out AI for precision cancer care
Source: University Hospitals
University Hospitals, located in Cleveland, Ohio, is using AI to deliver faster, more precise radiation treatment for cancer patients. The new system, called Varian Ethos 2.0, adapts to changes in the body before each session, improving accuracy and patient outcomes.
What’s happening: Ethos uses AI and high-performance computing to generate fresh scans of a patient’s body before every treatment. That allows physicians to update treatment plans on the spot, targeting tumors more precisely while avoiding healthy organs that may have shifted position since the last session.
Dr. Daniel Spratt, chair of radiation oncology at UH, says the system enables physicians to safely deliver higher doses of radiation, especially in hard-to-treat areas like the pancreas. This shortens treatment time and raises the likelihood of remission.
Pancreatic cancer is among the deadliest types of cancer, in part because the small intestine is so close to the pancreas. With older methods, there was no safe way to deliver strong doses of radiation in that area. Ethos changes that.
How it works:
Uses AI-powered imaging to map the body’s internal anatomy before every session
Updates the radiation plan in real time to match the patient’s anatomy
Continuously monitors breathing and body motion for precise beam placement
Mounted cameras track the skin and adjust for internal movement
Improves safety by avoiding healthy tissue and adapting to tumor shrinkage
So far, about 100 patients have received treatment using Ethos at UH Seidman Cancer Center. Eventually, the health system expects to treat more than 30 patients per day. The treatment is covered by commercial health insurance and requires no surgery or sedation.
University Hospitals is the only major hospital system in Northeast Ohio currently offering this AI-based radiotherapy. It’s part of a broader cancer care investment, including a planned $39 million regional cancer hub in Painesville set to consolidate UH’s cancer services.
Why it matters: AI is doing much more than helping with drug discovery and diagnostics. In this case, it’s actively improving how cancer is treated day by day. For patients, that means fewer sessions, better targeting, and a more comfortable experience.
🚀 Explore How Industries Are Leveraging AI + Orchestration
AI isn’t just hype — it’s driving real transformation across industries.
Curious how AI agents contribute to real-world results across finance, insurance, manufacturing, and more? Camunda’s latest AI infographics break it down with examples of autonomous agents and humans working together to streamline complex processes. Learn common use cases for deploying AI to enhance customer experience, employee productivity, and IT efficiency.
👉️ Explore the visuals to see what’s possible when AI meets intelligent automation.
🎇 Trump administration preps AI.gov platform for July 4 launch
Source: The White House
The federal government is building a comprehensive AI platform set to launch on Independence Day, according to leaked code from a GitHub repository that was quickly taken down after media inquiries.
The General Services Administration is developing AI.gov to "accelerate government innovation with AI" through three integrated tools—a chatbot, analytics dashboard, and API connecting to OpenAI, Google, Anthropic, Amazon's Bedrock, and Meta's LLaMA models.
Why it matters: The platform represents the Trump administration's most ambitious AI integration effort, spearheaded by former Tesla engineer Thomas Shedd, who leads GSA's Technology Transformation Services. Shedd previously told staff he wants to "AI-ify much of the government" through coding agents and contract analysis systems.
Yes, but: Government employees described internal reaction as "pretty unanimously negative," citing concerns about AI introducing security vulnerabilities, software bugs, and potentially recommending cancellation of critical contracts.
The initiative aligns with the Trump administration's broader AI push, originally championed by Elon Musk's now-concluded involvement with the Department of Government Efficiency, and new OMB guidance encouraging federal AI adoption while maintaining American-made preferences.
Whether agencies will be required to use the platform remains unclear, though Shedd has indicated mandatory adoption across government departments.
Want To Maximize Your Team’s Potential?
We all wish we could pull together our very own “Dream Team” at work – the only question is, how?
The answer is simple: Agentforce, SalesForce’s AI agent platform.
Unlock your team’s full potential with Agentforce:
Deploy proactive AI support
Agentforce offers always-on assistance for your employees or customers.
Customize its expertise
Equip it with your business knowledge for specialized, tailored help.
Automate key tasks
It autonomously executes tasks specific to its assigned role.
OpenAI’s teaming up with Google on cloud, rivalry aside
Glean’s now worth $7.2B thanks to the enterprise AI boom
Scammers are using AI to fake students and steal financial aid
Salesforce blocks AI rivals from using slack data
Silicon Valley led the quest for driverless cars. But Chinese robotaxis are catching up fast
XRobotics’ countertop robots are cooking up 25,000 pizzas a month
UK government harnesses Google’s AI to support faster planning decisions
🧠 Zuck’s building a team to chase down ‘superintelligence’
Source: ChatGPT 4o
Meta will spend $14.8 billion to acquire Scale AI and install CEO Alexandr Wang at a new 50‑person ‘superintelligence’ lab, a record outlay that founder‑CEO Mark Zuckerberg hopes will vault the company back into the front rank of AI research.
Bloomberg reports Zuckerberg has been hosting private recruiting sessions at his Lake Tahoe and Palo Alto homes, building a lab internally dubbed the "superintelligence group." Meta is finalizing a $14.8 billion investment in Scale AI to bring founder Alexandr Wang into the new lab.
Why it matters: The move represents Meta's largest external AI investment ever and signals Zuckerberg's personal acknowledgment that the company isn’t winning the AI race. Current and former Meta employees told CNBC that Zuckerberg has grown "agitated that rivals like OpenAI appear to be ahead in both underlying AI models and consumer-facing apps."
Meta: “We continue to make good progress on AI...” — Mark Zuckerberg, Q4 2024 earnings call
Optimist: DeepMind’s Demis Hassabis sees a 50 % shot at AGI within 5‑10 years.
Skeptic: Nature survey (Mar 2025) finds most researchers doubt current tech can reach human‑level reasoning.
Between the lines: Zuckerberg's hands-on involvement is unprecedented. He's reorganized Meta's Menlo Park office so the new team sits near him and is offering compensation packages between seven and nine figures (whoa) to poach talent from OpenAI and Google. This level of integration is signaling that Zuckerberg is back in “founder mode”.
Llama 4's April release was poorly received by developers, while the company's flagship "Behemoth" model has been delayed multiple times due to underperformance on reasoning and math benchmarks—exactly the capabilities where OpenAI and Google excel.
By the numbers:
Meta plans to spend up to $65 billion on AI infrastructure in 2025
Scale AI generated $870 million in revenue in 2024, expecting $2 billion this year
The company has laid off key AI researchers including FAIR leader Joelle Pineau
What's at stake: Scale AI represents the "third pillar" of AI development alongside chips and talent—high-quality training data. The startup uses PhD-level experts who craft medical, legal, and autonomous‑driving datasets—training material where accuracy is paramount.
Meta's superintelligence ambitions face significant technical hurdles. AGI requires models that can match human performance across all cognitive tasks—a capability researchers debate is years away or may have no clear path. Meta still lacks the specialized "reasoning" models that have become OpenAI and Google's competitive advantage.
The talent war is intensifying. A venture capitalist noted three instances last week of Meta losing AI talent to competitors offering over $2 million annually. Meanwhile, Zuckerberg has deprioritized Meta's research arm FAIR in favor of product-focused teams—a shift from long-term research to immediate competitive threats. Meta’s FAIR created PyTorch (2016) and open‑sourced Llama‑2 (2023), proving it can turn research into widely‑used tools
Meta's existing AI reaches 1 billion monthly users across its platforms, but Zuckerberg recognizes that consumer adoption won't matter if competitors achieve breakthrough capabilities first. The Scale AI deal represents his biggest bet that data quality and specialized talent can help Meta leapfrog rivals in the race to AGI.
Zuckerberg's superintelligence gambit represents the purest form of founder mode we've seen in Big Tech—a CEO acknowledging that his company's survival depends on personally driving an existential priority rather than trusting traditional corporate processes.
When a founder hosts recruiting dinners at his personal homes and rearranges office furniture to sit closer to critical teams, it signals that normal management structures have broken down.
Zuckerberg has done this twice before during existential threats—pivoting to mobile in 2012 and to video/Stories to counter Snapchat. Both times required him to override internal resistance and corporate inertia. The AI race presents a similar inflection point, but with higher stakes.
The $14.8 billion Scale AI bet is about acknowledging that Meta's internal AI development has hit a wall. When you're spending more on a single external investment than most companies' entire market cap, you're not optimizing—you're panic-buying.
Zuckerberg's track record suggests he excels at platform shifts, but AGI requires scientific breakthroughs, not just product execution. No amount of personal involvement can compress the timeline for fundamental research, and in AI, being second often means being irrelevant.
Meta is betting close to a quarter of its 2024 cash flow that better data and hand‑picked talent can close the AGI gap.
Which image is real? |
🤔 Your thought process:
Selected Image 1 (Left):
“AI loves neon pink and purple.”
“The other image looked too futuristic to be real. Too much blending into the background”
Selected Image 2 (Right):
“The other image doesn't have enough detail - that fooled me!”
“I just thought it looked cool.”
💭 A poll before you go
Do you think Meta is losing the AI race? |
Thanks for reading today’s edition of The Deep View!
We’ll see you in the next one.
P.S. Enjoyed reading? Take The Deep View with you on the go! We’ve got exclusive, in-depth interviews for you on The Deep View: Conversations podcast every Tuesday morning. Subscribe here!
If you want to get in front of an audience of 450,000+ developers, business leaders and tech enthusiasts, get in touch with us here.