Happy Thursday. It's April 2nd, 2026, and here's what's moving today. Eli Lilly just signed the largest AI drug discovery deal in history — $2.75 billion with Insilico Medicine. Bitcoin is whipsawing around $66,000 as geopolitical noise from Iran and surging oil prices rattle risk assets. Metaplanet quietly jumped to third place among the world's largest Bitcoin treasury companies after scooping up over 5,000 BTC in Q1. And Tether is finally submitting to a full independent audit by KPMG, while its new USAT stablecoin expands beyond Ethereum onto Celo. Let's get into it.
Let's start with what might be the clearest signal yet that AI-driven drug discovery has crossed from experimental curiosity into big pharma's core strategy. Eli Lilly has signed a $2.75 billion deal with Insilico Medicine — the largest AI drug discovery partnership ever. The structure: $115 million upfront, with the rest in milestone payments and royalties tied to actual drug development, manufacturing, and commercialization.
What Lilly gets is exclusive global rights to Insilico's Pharma.AI platform. This thing uses generative AI and large language models to identify disease targets, design novel molecules, and predict clinical outcomes. Insilico isn't some startup pitching slides — they've already pushed 28 AI-identified drug candidates through their pipeline, with nearly half at the clinical stage. That's real output.
And Lilly isn't putting all its AI eggs in one basket. They've also committed $1 billion with Nvidia to build out broader AI capabilities. Meanwhile, Roche is deploying advanced AI hardware across its own research operations. The message from big pharma is unmistakable: AI isn't a nice-to-have anymore, it's becoming the backbone of how drugs get discovered and developed.
What makes this deal particularly significant is that it's not just about speeding up research — it's about commercial viability. Lilly is integrating AI-designed compounds directly into its development pipeline. We're past the proof-of-concept phase. This is AI producing drugs that a $700 billion pharmaceutical company is willing to bet billions will actually reach patients. The timeline compression alone could be worth the investment — traditional drug discovery takes 10 to 15 years and costs upward of $2 billion per approved drug. If AI can meaningfully cut either number, the economics shift dramatically for the entire industry.
The Bitcoin mining industry is going through a genuine identity crisis right now. CoinShares just released its latest mining report covering Q4 2025, and the numbers are rough. Hash price — the revenue miners earn per unit of computational power — dropped below $30 per petahash per day. That's a 5-year low. About 15 to 20% of older generation mining machines are now unprofitable. The network hash rate pulled back roughly 10% from its peak.
And then there's the exodus. Bitfarms, one of Canada's largest miners, is rebranding entirely. They're becoming Keel Infrastructure, relocating to New York, and pivoting to AI and high-performance computing. They plan to sell their entire Bitcoin stash — around 2,400 coins worth about $161 million — to fund AI data center projects. They're not alone. Over $70 billion in AI-related infrastructure contracts have been announced across the mining industry. Soluna just bought a $53 million wind farm specifically to power an AI facility. The economic logic is straightforward: AI workloads offer higher margins and more predictable revenue than Bitcoin mining.
On chain, Riot Platforms moved 500 BTC — roughly $34 million — off their wallets. Genius Group liquidated its entire Bitcoin treasury to pay down $8.5 million in debt. The treasury selloff pattern is real and growing.
But here's the contrarian take. CoinShares expects hash rates to rebound to 1.8 zettahashes per second by end of 2026. The miners who survive — the ones with low costs, low leverage, and modern hardware — will be positioned in a much less crowded field. Companies like HIVE and CleanSpark are running lean balance sheets with cost advantages. If Bitcoin recovers above $100,000, the survivors inherit a dramatically better competitive landscape. The shakeout is painful, but it's also exactly how Bitcoin mining has always worked: the halving squeezes out the weak hands, and the network emerges more efficient on the other side.
Meanwhile, the energy convergence between Bitcoin mining and AI is increasingly real. Both industries need the same thing — massive, reliable, often stranded power sources. Whether it's natural gas, nuclear, or renewables, the physical infrastructure is the same engine. The miners who figure out how to serve both markets may end up more resilient than either pure-play miners or pure-play AI data centers.
Okay, let's zoom out to the institutional picture. March was actually a decent month for Bitcoin ETFs. Net inflows came in at $1.32 billion, breaking a 4-month streak of outflows. BlackRock led the charge with nearly $98.4 million flowing into IBIT. And last week alone, BlackRock and Fidelity collectively purchased about $400 million in Bitcoin, while selling around $250 million — so net positive accumulation even during a period of market stress.
BlackRock is also pushing forward with a new product: a Bitcoin Premium Income ETF that would generate yield by writing covered call options against its existing IBIT holdings. That $55 billion in IBIT assets could become the base for a yield-generating instrument. Think about what that means — it's not just about Bitcoin exposure anymore, it's about Bitcoin as a productive financial asset within traditional portfolio construction. If this gets SEC approval, it could pull in a whole new class of institutional money that wants Bitcoin exposure but also wants income.
But the picture isn't all rosy. Early April saw $173.8 million in outflows from spot Bitcoin ETFs, and retail demand remains notably weak. The Coinbase Premium Index — a rough gauge of U.S. retail buying pressure — has been declining. So what we're seeing is a bifurcated market: institutions are methodically accumulating on dips while retail sits on the sidelines.
Add to that some important infrastructure moves. Coinbase just received conditional approval from the OCC for a national trust charter. If finalized, this makes Coinbase a federally regulated crypto custodian — a big deal for institutional clients who need that regulatory wrapper. And SoFi launched a 24/7 banking hub that lets companies hold dollars, convert to stablecoins, and move money instantly within a regulated bank. These are the plumbing upgrades that make institutional adoption sticky and permanent.
Grayscale's assessment is probably the most balanced: investors are sidelined by Middle East tensions and the oil shock, but the structural adoption trends haven't reversed. The question isn't whether institutions want Bitcoin — it's when the macro noise clears enough for them to deploy more aggressively.
Let's close with stablecoins, because two things happened this week that could reshape the competitive landscape.
First, Tether is finally getting a real audit. KPMG has been confirmed as the auditor, with PwC assisting on internal systems preparation. For years, Tether resisted independent scrutiny, relying instead on quarterly attestation reports. The motivation here seems partly commercial — reports suggest Tether's attempt to raise $20 billion from investors fell short, and they've scaled back to a $5 billion target. Nothing focuses the mind on transparency like needing to convince sophisticated investors to hand over capital. Now, KPMG's track record isn't spotless — they were the auditor for Silicon Valley Bank and Signature Bank, both of which failed — so skepticism about the audit's rigor is fair. But the mere fact that Tether is submitting to this process is significant. It changes the dynamics with regulators and institutional counterparties.
Second, Tether's newer USAT stablecoin — the one designed specifically for U.S. compliance under the proposed GENIUS Act — just expanded beyond Ethereum onto the Celo blockchain. This is its first deployment on another network. Why Celo? It's mobile-first, low-cost, and already has 14 million Opera MiniPay wallet users. Google Cloud is backing the distribution infrastructure. USAT is a different beast from USDT — it has full reserve backing managed by Anchorage Digital with regular attestations by Deloitte. It's essentially Tether's answer to Circle's USDC in the compliance-first stablecoin market.
Speaking of the GENIUS Act, the U.S. Treasury just published its first proposed rulemaking under that legislation. The text builds out the operational architecture for stablecoin governance: which institutions can issue payment stablecoins, under what conditions, and at what scale before federal oversight kicks in. This is the regulatory framework that will determine who wins the stablecoin race. Tether is clearly trying to position itself on both sides — maintaining USDT's dominance in global markets while building USAT to meet U.S. compliance requirements head-on.
Meanwhile, Circle's response time during the $280 million Drift Protocol exploit on Solana drew sharp criticism. Stolen USDC reportedly moved for hours without being frozen. If stablecoins are going to serve as critical financial infrastructure, the expectations around operational response times are only going to intensify.
Here's the thread connecting everything today. Whether it's Eli Lilly betting $2.75 billion that AI can design real drugs, or Bitcoin miners pivoting their power infrastructure toward AI workloads, or BlackRock packaging Bitcoin into yield products — the pattern is the same. The experiments are over. The question now is execution, and who builds the infrastructure that actually scales. Keep your eyes on the builders, not the noise. That's it for today. I'll talk to you tomorrow.