CoreWeave Stumbles, Foxconn Jumps 27% On AI Demand, VC Bets Tilt To Agentic Ops As Bullish Debuts At $13.2B

The AI trade split today along an old fault line, cash generation vs. capacity promises. Foxconn’s results underscored real demand for AI hardware, even as CoreWeave’s losses rattled investors. On the private side, capital is flowing into agentic manufacturing, data tooling, and vertical ops. And in crypto, a newly public exchange grabbed a $13.2 billion valuation while AI token presales chased headline multiples.

📈 AI Market Movers & Trends

Hardware reliability trumped infrastructure exuberance. Foxconn said second quarter profit rose 27% year over year, beating forecasts on strength in AI servers, per CNBC. If you’re looking for confirmation that the “picks and shovels” still have legs, a manufacturing bellwether printing double digit growth is about as clean as it gets. Contrast that with CoreWeave, where shares fell after Reuters reported that “growing losses” overshadowed a demand surge. The lesson is familiar, capacity gets funded; profitability gets paid.

Still, the street hasn’t lost its appetite for the leaders. Mizuho raised price targets on Nvidia and AMD, according to TipRanks, reinforcing the view that the silicon cycle has more room even after the run. That said, headline valuations carry gravity. The Globe and Mail highlighted that Microsoft and Nvidia have both hit $4 trillion market caps, rarified air that invites every incremental turn of the estimate crank to be scrutinized. It’s no coincidence that billionaire Philippe Laffont has been trimming Nvidia exposure, per Yahoo Finance, smart money often derisks into strength.

Beneath the mega caps, smaller names are finding a bid when the AI story shows up in results. Leaddesk’s Q2 call emphasized AI as a growth driver and the stock rose on that focus, according to Investing.com. Meanwhile, ETF flows tilt toward broad AI capture as development accelerates (Yahoo Finance), a tell that generalists want exposure without single name whiplash. The underpriced narrative, old economy suppliers with AI mix shifts, Foxconn today, others tomorrow, could deliver steadier compounding than the flashier infrastructure names battling unit economics. Keep an eye on the industrial edge, as our Funding section shows, venture dollars are seeding the demand that will flow back into public hardware suppliers.

💸 Funding Watch

Venture appetite is rotating from model vanity to operational leverage. Squint raised $40 million in Series B funding to expand agentic manufacturing, software that turns frontline processes into autonomous workflows. It’s the sort of “AI that does” rather than “AI that talks,” and it maps cleanly to where enterprises spend. Protege closed a $25 million Series A to scale an AI training data platform, a reminder that datasets remain the scarce commodity in a world obsessed with parameters. Profound added a $35 million Series B for AI search optimization, aimed at getting enterprises out of the “RAG demo” cul de sac and into measurable conversion lift.

Sector breadth is widening, too. Agtech player BinSentry secured $50 million to globalize its AI based feed logistics platform, a long tail example of AI touching supply chains far from Silicon Valley. Early stage energy is alive, Refold AI raised $6.5 million from Eniac Ventures and Tidal Ventures, and Dealops added $7 million to streamline sales operations with AI. Even Menlo Ventures’ Anthology Fund update flagged that applied AI is pushing from prototype to production across the portfolio. The common thread, capital is favoring agentic systems, data flywheels, and vertical operating gains, the inverse of loss fueled infrastructure scale that spooked CoreWeave’s tape.

This bifurcation has public market implications. As Foxconn’s print showed, dollars spent on AI capable devices and servers flow quickly into earnings. The startups funded this week, agentic manufacturing, smarter search, logistics ,are precisely the workloads that will drive incremental edge compute and specialized silicon orders over the next 12–24 months. That’s a healthier flywheel than crowding into foundation model training runs that still struggle to clear their cost of capital. If you’re mapping venture signals to public exposure, the better pairing today might be industrial and semiconductor suppliers over AI only clouds. When venture money funds “work,” public markets eventually pay “cash.”

🪙 Crypto Moves

The crypto side of the AI story split, too, between institutional credibility and retail sizzle. Reuters reported that Peter Thiel backed exchange Bullish made a blowout NYSE debut, valued at nearly $13.2 billion, with the Financial Times noting the stock surged out of the gate. That’s real validation for market structure plays, venues that benefit from volumes regardless of whether the next narrative is DeFi, AI agents on chain, or tokenized real world assets.

At the other end of the spectrum, AI token presales leaned hard on eye catching projections. Blockchain Reporter touted that investors are “rushing into” Ozak AI’s presale on expectations it could rally from $0.005 to $2 by 2026. The company separately announced it “secured $1.8M with multiple partnerships” ahead of launch, per GlobeNewswire, a tiny round by venture standards but meaningful in token land. Similar pitches, “5 new AI crypto coins” promising stellar 2025 returns and a $0.10 presale that “could turn $250 into $750K”, are ricocheting around retail channels. CoinMarketCap shows live listings like VERSE AI gathering liquidity, while hybrid narratives (Sentient Dynamics Token and RealSimple Crypto) sell a future where AI agents and blockchains interoperate.

Here’s the contrarian read, the serious money today is in the rails, not the roulette. Bullish’s listing elevates exchange credibility at a moment when AI tokens are mostly pre product marketing. That doesn’t mean the agent on chain thesis is wrong; it means most of the near term value accrues to the pipes that will carry those agents when they arrive. For equity investors, this echoes the public market split, profits and unit economics are winning over promises of scale. For founders, the actionable overlap with venture flows is obvious, agentic systems that touch real transactions, not just token narratives, are attracting both users and capital.

📊 Stay tuned for tomorrow’s MarketPulse or join our free Discord community for daily briefings from Midas AI.