📍 Venture capital wrote roughly $798.5 million in checks across ten rounds on June 17 alone, and every dollar chased the same thing — the infrastructure that makes AI usable inside real institutions, from Twenty's $1 billion cyber-warfare valuation to Odyssey's $1.45 billion world-model bet. On the same day, Crunchbase profiled Black founders who raised nearly $30 million, exited, and crossed to the other side of the table to write checks themselves — betting that AI has "brought down the walls" of building a company. When the money moves this fast toward who controls the machines, the only question that matters is whether you're positioned to own a piece of the layer everyone is now paying for. ───── Hey Builders! The headline number from yesterday is $798.5 million across ten venture rounds, but the number that tells the real story is zero. That is roughly how much of that capital flowed to consumer-facing AI applications. The money went somewhere more durable — into the control layer that sits between frontier models and the institutions that have to trust them. Tech Startups framed it precisely in its June 17 roundup. "Venture is still willing to fund ambition, but it is increasingly paying for the removal of constraints," the report concluded. "The best-funded startups today are not merely generating intelligence. They are making intelligence deployable." That is the thesis of the entire week. Odyssey raised $310 million to simulate the physical world. Twenty raised $100 million to industrialize offensive cyber operations. Convey raised $38 million to run enterprise operations autonomously. Behavox, NeuralTrust, Trace Finance, and Clario all raised to govern, secure, or clean the systems around AI rather than to build the models themselves. Step back further and the concentration is staggering. Crunchbase reports global venture funding hit a record $300 billion in Q1 2026, with AI capturing 80% of the total and just four giant deals accounting for 65% of all global venture funding in the quarter. CB Insights found that $100 million-plus AI rounds now make up 94% of total AI funding value. Capital is not spreading across the market. It is clustering into a handful of companies that own the scarcest pieces of the stack. Against that backdrop, a more interesting story than the funding gap is what builders are doing about it. Crunchbase profiled founders like Clarence Bethea, who raised nearly $30 million for warranty startup Upsie and sold it to Akko in 2024, and Cortney Woodruff, who built two venture-backed companies — both of whom crossed to the other side of the table to write checks and rewrite the rules from inside. Their read on the moment is not defeat but leverage. "AI brings down the walls of building an MVP, talking to customers, and starting to gain traction," Bethea told Crunchbase. The same concentration squeezing the market is also handing outsiders the cheapest tools in history to build with. Microsoft made the week's other defining move. At Build 2026 on June 2, the company launched seven in-house AI models, led by MAI-Thinking-1, its first reasoning model — trained from scratch on commercially licensed data with no distillation from OpenAI. The strategic message was unmistakable. Even Microsoft, OpenAI's largest backer, has decided it cannot afford to depend on a single supplier for the capability that now defines its business. That is the unifying lesson for builders this week. Whether you are a hyperscaler or a seed-stage founder, the people writing and earning the biggest checks are the ones who own a defensible layer — the model, the compute, the controls, the rails — rather than renting it. Ownership is no longer a values statement. It is the entire game. ───── 1️⃣ Twenty Hits a $1 Billion Valuation on a $100 Million Series B to Industrialize Cyber Warfare Arlington-based Twenty announced a $100 million Series B on June 17 at a $1 billion valuation, becoming what the company describes as America's first VC-backed cyber-warfare startup, per PR Newswire and SiliconANGLE. Accel led the round, with Friends & Family Capital, Point72 Ventures, and Caffeinated Capital participating. The financing brings Twenty's total raised to $138 million. Founded in 2024, Twenty builds AI-enabled, end-to-end offensive cyber systems for the U.S. military and intelligence community, designed to give operators the speed and scale to deter and defeat adversaries in cyberspace. The company says it keeps human judgment in the loop, a distinction that matters as much for federal procurement credibility as for ethics. Its earlier backers include General Catalyst and In-Q-Tel, the CIA's venture arm, which signals that capital once reserved for traditional defense channels is now flowing through venture rails. Twenty's raise is the clearest defense-tech signal of the week, and it reflects a broader shift. Investors who spent years backing drones and hardware are now moving toward software-defined capabilities with direct mission relevance. The round values a two-year-old company at a billion dollars precisely because it sits at the intersection of national security and AI, a part of the stack where failure is expensive and switching costs are permanent. 💡 For Founders Defense and public-sector demand is one of the few markets where AI software commands premium valuations on day one, because the buyer cannot tolerate failure and rarely switches. If you are building anything adjacent to national security, regulated infrastructure, or government workflows, structure your company for procurement credibility early — human-in-the-loop design, security clearances, and mission relevance are now fundable assets, not afterthoughts. ───── 2️⃣ Odyssey Raises $310 Million at a $1.45 Billion Valuation to Build AI That Simulates the Physical World Odyssey closed a $310 million Series B on June 17, the largest startup round of the day, at a $1.45 billion valuation, per Tech Startups and Reuters. Natural Capital led the round, with Amazon, AMD Ventures, GV, EQT, and In-Q-Tel joining alongside existing backers including Elad Gil, Garry Tan, Guillermo Rauch, and Kyle Vogt. The Palo Alto company also named AWS its preferred cloud provider, a reminder that compute partnerships are now woven into the financing itself. Odyssey is building "world models" — systems that learn to predict and interact with the physical world over long horizons. Unlike a language model that manipulates text, a world model learns the physics of reality, which makes it foundational technology for robotics, autonomous systems, defense, gaming, and scientific simulation at once. CB Insights has identified physical AI as one of the fastest-rising themes of 2026. The investor list tells the strategic story. When a chip investor like AMD Ventures, a hyperscaler like Amazon, and a national-security fund like In-Q-Tel all show up on the same cap table, they are not betting on a product. They are betting on infrastructure that could spill into a dozen markets. If large language models taught investors to pay for intelligence, world models are teaching them to pay for embodied prediction. 💡 For Founders Watch where physical AI is heading, because the next wave of buildable industries — robotics, autonomous logistics, industrial simulation — will run on top of world models the way today's apps run on top of LLMs. If you are early, you do not need to build the model; you need to identify the vertical application that becomes possible once embodied prediction gets cheap, and position to own that workflow before incumbents notice. ───── 3️⃣ Microsoft Launches Seven In-House AI Models, Cutting Its Dependence on OpenAI At Build 2026 on June 2, Microsoft unveiled a family of seven homegrown AI models led by MAI-Thinking-1, its first reasoning model, per Microsoft AI, CNBC, and Windows Central. MAI-Thinking-1 carries 35 billion active parameters in a sparse Mixture-of-Experts architecture with roughly one trillion total parameters and a 256,000-token context window, and Microsoft says it was trained from scratch on commercially licensed data with no distillation from third-party models such as OpenAI's GPT series. The seven models span reasoning, coding through MAI-Code-1-Flash, image generation, transcription, and voice. In blind evaluations run by Microsoft's independent human-rating partner Surge, the company says MAI-Thinking-1 was preferred over Anthropic's Claude Sonnet 4.6 and matches Claude on coding benchmarks. Microsoft framed the effort as building "a hill-climbing machine" — an internal capability it can improve on its own timeline. The relationship with OpenAI has not been severed. Azure remains OpenAI's primary infrastructure, and GitHub Copilot and Microsoft 365 Copilot still run OpenAI models. What changed is the dependency. Microsoft has decided that the future of Azure AI will not be built around any single outside supplier, no matter how important. The largest backer of the most valuable AI company on earth just built its own escape hatch. 💡 For Founders If Microsoft will not stake its future on one model provider, you should not either. Audit every part of your product that depends on a single vendor's API, and build the optionality to swap models, providers, or compute without rewriting your business. The companies that own their critical dependencies — or can route around them — will keep their margins when pricing power shifts. ───── 4️⃣ Black Founders Who Raised Nearly $30 Million Are Crossing the Table to Become the Investors A generation of Black founders who fought to raise capital is now crossing to the other side of the table to write the checks, Crunchbase reported on June 17. Clarence Bethea raised nearly $30 million for his extended-warranty startup Upsie before selling it to Akko in 2024, then joined his backer True Ventures as an investor in 2023. Cortney Woodruff built and raised venture capital for two companies, including the learning platform Assemble he co-founded with actor Jesse Williams, before turning to angel investing. Both moved into capital allocation for the same reason — to change the dynamics they faced as founders. What they found on the other side reframed the problem from personal to structural, and from hopeless to fixable. "What surprised me was how much venture capital is driven by pattern recognition," Woodruff said, describing how investors source from familiar circles and back founders who feel familiar to them. His conclusion is not that the economics of venture must change, but that the people doing the recognizing must. "The most successful investors in the future may be the ones who can recognize extraordinary opportunities in places others have been trained to overlook." Both founders are pointing at the same lever — AI — as the equalizer. Bethea built his current venture, an unfiltered fundraising-education platform called "What VCs Won't Say," on more than 75 videos and 90 workbook pages, and says its frameworks have already helped two founders raise millions. He is, by his own account, "more optimistic than ever before," because technology is collapsing the cost of getting from idea to traction. That optimism is the story the funding tables miss. 💡 For Founders The fastest way to change a system that runs on pattern recognition is to put more of your own people in the rooms where patterns get set — so if you exit or gain traction, consider angel investing and mentorship as a second act, the way Bethea and Woodruff did. And use the equalizer they named: AI has collapsed the cost of building an MVP, reaching customers, and proving traction, so build the evidence that makes you impossible to overlook before you ever walk into a pitch. ───── 5️⃣ Convey Raises $38 Million Series A to Run Enterprise Operations With AI Teammates Convey announced a $38 million Series A on June 17, led by Andreessen Horowitz with continued backing from Khosla Ventures and Pear VC, per Tech Startups and Business Insider. The San Francisco company sells "AI teammates" that let non-technical operators build and manage autonomous workflows, and it says it has already completed more than one million hours of automated work. Convey's customer list already includes NBCUniversal, Samsara, TelevisaUnivision, Unity, Faire, and ChargePoint — enterprise logos that signal the product is running in production, not pilots. What distinguishes the round is positioning. Business Insider's reporting highlighted Convey's pitch as outcome-oriented rather than task-oriented, which is exactly what enterprise buyers now demand from a crowded field of agent startups. The lesson sits at the center of this week's thesis. Investors are no longer rewarding generic AI assistants bolted onto chat interfaces. They are paying for vendors that own a specific operational loop and can prove they remove labor, latency, or cost from it. Orchestration with real workflow ownership is the defensible position; thin wrappers around general-purpose models are the margin trap. 💡 For Founders If you are building with AI agents, stop selling tasks and start owning outcomes — enterprise buyers reward the vendor that takes a full operational loop off their plate, not the one that generates content faster. Anchor your pitch in measurable work completed, like Convey's million-plus automated hours, because production traction inside real enterprises is now the metric that separates a strategic infrastructure bet from another app. ───── 🔧 Three moves to make this week 1️⃣ Audit your single points of dependency Microsoft just built seven models to avoid depending on one supplier, and you should map your own version of that risk this week. List every vendor, API, or platform your product cannot function without, and identify at least one swappable alternative for each — dependency you cannot route around is leverage someone else holds over your margins. 2️⃣ Treat your network as fundable infrastructure With fewer deals closing and relationships deciding who reaches the top of the investor stack, schedule three conversations this week with operators or adviser-investors in your sector — not to pitch, but to build. Attend the industry-specific conference months before you need to raise, because the warm introduction you build now is the round you close later. 3️⃣ Reposition your AI product around owned outcomes If any part of your offering is a thin wrapper on a general-purpose model, rewrite your pitch this week to center on a specific operational loop you own end-to-end. Quantify the labor, cost, or latency you remove, and target the regulated or mission-critical workflows where buyers tolerate real spend because failure is expensive. ───── 💬 Quote of the Day "AI brings down the walls of building an MVP, talking to customers, and starting to gain traction. That's really exciting for founders who don't fit the normal founder stereotype." — Clarence Bethea, founder of Upsie, now managing partner of What VCs Won't Say ───── 🏁 Stay Ahead With OHUBNext This brief is part of what OHUBNext members get every day. For $5.99/month — or $59/year — you get the full daily brief, access to OHUB's Library of Opportunity, self-paced certificates in High-Growth Company Building and Tech Ecosystem Investing, career accelerator tools, and live monthly labs with the OHUB team. The annual plan includes exclusive masterclasses, early course access, and wealth tools built for builders who are serious about owning what comes next. Institutional-grade. Fraction of the cost. 🚀 Join at opportunityhub.co/next ───── 🎬 Closing Thought When venture capital writes nearly $800 million in a single day and sends every dollar toward the control layer, when Microsoft builds its own models rather than depend on its closest partner, and when founders who fought to raise capital cross the table to start writing the checks, a single pattern comes into focus. The AI economy is concentrating — into the hands of those who own the model, the compute, the controls, and the rails. The opening is that the people deciding who owns those layers are no longer a fixed set. That concentration is the threat and the opportunity at once. The same forces routing $300 billion through four deals a quarter are also defining, in real time, which layers of the stack will hold value for the next decade. Odyssey's world models, Twenty's cyber systems, Convey's owned workflows — these are not products chasing a trend. They are claims staked on scarce ground while the ground is still being mapped. For founders building outside the established networks, the data is honest about the difficulty and clear about the path. Capital flows to who you know and to what you own. Both can be built — deliberately, starting this week — but neither will be handed over. The window where these layers are still contestable is open now and closing fast. The builders who understand that ownership is the entire game won't just survive the concentration. They'll own a piece of the layer everyone else is renting. ───── ⚡️ OHUBNext Daily Brief - investments, edge tech, and moves that matter. For 12+ years, OHUB has been building pathways and on-ramps to multi-generational wealth without reliance on pre-existing wealth. Through exposure, skills, entrepreneurship, capital markets, and inclusive ecosystems, we've helped people create new jobs, new companies, and new wealth.