The chart doesn't lie, but the narrative often does. In late 2025, a relatively low-profile Pre-IPO fund, MSX, announced its third tranche with a portfolio that reads like a science fiction novel's table of contents: Neuralink and Anduril. On-chain data doesn't capture this, but the capital flow does. The ledger of private markets shows a coordinated bet on the two most extreme applications of artificial intelligence: the invasion of the human brain and the automation of the battlefield.
This is not a diversified tech fund. It is a concentrated wager on the thesis that AI's next trillion dollars of value creation will not come from better chatbots or faster GPUs, but from physical and biological convergence. MSX is not buying into revenue. It is buying into a narrative with a hard technological edge.
The announcement was thin. Two names, a few tags, a suggestion of 'AI's next frontier.' That is it. No financial details, no valuation metrics, no exit timeline. But for those of us who have spent the last decade dissecting smart contracts and capital flows, the message is clear. MSX is positioning itself at the IPO window for two of the most anticipated technology debuts of the decade. The question is whether the underlying engineering can support the valuation narrative.
This analysis will cut through the press releases. We will examine the technical maturity of both companies, their commercialization paths, and the brutal math of their current valuations. Based on my audit experience across 45,000 lines of smart contract code and my forensic analysis of the Terra/Luna collapse, I have learned that the process reliability of the underlying technology is the only thing that matters when the hype fades. Smart contracts have no mercy, and neither does the market when a POC fails to scale.
Context: The Pre-IPO Playbook and the 'AI + Hard Tech' Thesis
To understand the MSX move, you have to understand the Pre-IPO fund playbook. These vehicles typically enter 1-2 years before a company's public listing, providing late-stage liquidity to founders and early investors in exchange for a slice of the upside. The target return is often 3-5x, and the holding period is usually 2-3 years. The entire model depends on a clear, unimpeded path to an IPO.
MSX's third tranche fits this model perfectly. Neuralink, currently in the FDA IDE clinical trial phase, has no commercial revenue. Its estimated valuation is $8-10 billion, based on the last funding round and market chatter. Anduril, on the other hand, is a revenue-generating machine, projecting $1 billion in sales for 2025, a 150-180% year-over-year increase. Its valuation is estimated between $25-30 billion. Both are in the sweet spot for a Pre-IPO fund.
The 'AI + Hard Tech' label is accurate but insufficient. This is about specific sub-sectors: brain-computer interfaces (BCI) and defense AI. These are not adjacent to the current AI boom; they are the logical endgame of it. If AI is the new electricity, Neuralink is the wiring for the human nervous system, and Anduril is the circuit breaker for national security. MSX is building a portfolio that spans the biological and physical extremes of the AI revolution.
The timing is also critical. Neuralink is expected to IPO in 2027-2028, contingent on FDA approval for commercial sales. Anduril is expected to go public in 2026-2027. The MSX fund's typical 2-3 year exit horizon aligns perfectly with these windows, provided the FDA cooperates and the Pentagon's budget holds.
But the market is a harsh auditor. Follow the TVL, not the tweets. In the private markets, TVL is replaced by the backlog of contracts and the depth of the clinical trial pipeline. Let's examine the balance sheets of these two narratives.
Core Analysis: The Technical and Commercial Evidence Chain
Neuralink: The POC-to-Production Chasm
Neuralink's technical progress is real but painfully slow. The N1 implant, with its 1,024 flexible electrodes, is a marvel of engineering. The R1 surgical robot, designed to insert these threads while avoiding blood vessels, is equally impressive. As of late 2025, the company has three human patients implanted. The FDA granted Breakthrough Device Designation in 2025, which accelerates the review process but does not guarantee approval.
The most recent milestone is the 'CONVOY' feasibility study, approved in November 2025, which links the N1 chip to a robotic arm, allowing patients to control it via thought. This is a significant step towards the company's stated goal of restoring autonomy to paralyzed patients. But it is still a POC. The sample size is three. The long-term safety data on electrode degradation and signal attenuation is nonexistent.
In my forensic analysis of the Terra/Luna collapse, I mapped the exact block height where solvency failed. For Neuralink, the equivalent failure point is the FDA's PMA (Pre-Market Approval) process. The IDE phase is designed to catch catastrophic failures. The transition from IDE to PMA requires thousands of patient-hours of safety data. Neuralink has a handful. The gap between a successful POC and a scalable, commercial product is a graveyard of good ideas.
From a financial engineering perspective, Neuralink's valuation of $8-10 billion is based on a price-to-research ratio of 50-100x, given their ~$100 million annual R&D spend. This is not a company; it is an option on a future technology. The premium is justified only if the technology achieves a 30-50% market share in a BCI market projected to reach $10 billion by 2030. That is a massive 'if'. The competition is not idle. Synchron, with its non-invasive Stentrode, has a safer profile and easier regulatory path, albeit with only 16 channels. Precision Neuroscience is also making strides with its minimally invasive approach.
Neuralink's competitive advantage is channel count and bandwidth. The question is whether the market values bandwidth over safety. History suggests the market penalizes risk, not complexity. The ledger remembers everything, and the FDA's adverse event log will be the ultimate arbiter.
Anduril: The Scale-Up Machine
Anduril is a different beast. The Lattice software platform is the core of its 'software-defined defense' model. It integrates sensors, data, and autonomous systems into a single decision-making loop. The product portfolio spans the Ghost 4 drone, the Dive-LD autonomous submersible, and the Roadrunner missile system. The December 2025 acquisition of the UK-based 3D printing drone company Callum for $1.3-1.4 billion is a strategic move to bring hardware manufacturing in-house, addressing a key weakness in its 'software-first' approach.
The revenue numbers are solid. $1 billion in 2025, with a projected $2.5 billion contract from the Defense Innovation Unit (DIU) for 'large-scale expeditionary warfare' systems. The partnership with OpenAI in June 2025 to develop counter-UAS systems is a powerful endorsement of the technology's maturity. Anduril is not a POC; it is a scale-up. The contracts are signed, the hardware is deployed, and the revenue is flowing.
But the valuation is where the discipline comes in. At a $25-30 billion valuation, Anduril trades at a P/S ratio of 25-30x. Palantir, its closest public market comp, trades at a similar multiple. The defense AI market is projected to be worth $2.5 trillion globally, with Anduril targeting a 10-20% share. The addressable market is real, and the growth rate is exceptional. The risk is concentration. The US Department of Defense is a single customer. If the political winds shift and the defense budget is cut, or if a traditional prime contractor like Lockheed Martin successfully integrates AI into its own platforms, Anduril's growth story gets a severe haircut.
The acquisition of Callum is a double-edged sword. It improves vertical integration but also brings the complexities of hardware manufacturing, a low-margin, capital-intensive business that is foreign to a pure-play software company. The 'asset-heavy' pivot could dilute the gross margins that justify its high P/S multiple.
The MSX Thesis: A Portfolio of Extremes
MSX is betting on the convergence of these two trajectories. The portfolio is designed to capture the value of AI as it moves from the digital realm to the physical and biological worlds. It is a hedge on the future of human-machine interaction and autonomous national defense. The fund's logic is straightforward: invest in the leaders of these sub-sectors before the public markets assign a fair value.
But the data reveals a critical asymmetry. Anduril is a revenue-generating company with a clear path to profitability (expected by 2026). Neuralink is a pre-revenue experiment with a binary outcome: FDA approval or failure. MSX's 3-5x return target is heavily dependent on Neuralink's regulatory success. If Neuralink's clinical trials hit a snag, the entire portfolio's performance is compromised. The correlation between the two assets is low, but the systemic risk is high. A major adverse event in Neuralink's trial could spark a sector-wide sell-off in BCI-related assets, dragging down Anduril's narrative-driven valuation.
The 'From Brain to Battlefield' narrative is compelling, but it masks a fundamental difference in technological maturity. Anduril is a proven system; Neuralink is a promising prototype. Pairing them in a single fund suggests a belief that the market will reward the narrative as much as the technology. In my experience, that is a dangerous assumption.
Contrarian Angle: Correlation Does Not Equal Causation
There is a seductive logic to the MSX portfolio: 'AI is eating the world, so invest in the two most extreme use cases.' But this is a narrative fallacy. The correlation between Neuralink's success and Anduril's success is approximately zero. One is a medical device company navigating a decades-old regulatory framework; the other is a defense contractor navigating a politically-driven procurement process. The only commonality is the 'AI' label.
Investors are paying for the story, not the underlying data. Neuralink's valuation is a pure bet on regulatory approval and clinical scalability. Anduril's valuation is a bet on geopolitical instability and the continued shift towards autonomous warfare. These are different bets, and the risk profiles are not complementary. In a diversified portfolio, you would want these to be separate positions, not a single concentrated bet.
The blind spot here is the assumption that being first to market with a superior technology guarantees market leadership. The ledger of tech history is filled with companies that had the best product but lost the market to a competitor with a better go-to-market strategy or a more favorable regulatory environment. Synchron, with its less invasive approach, could easily win the BCI market not because it is better, but because it is safer and easier to approve. Similarly, Palantir, with its deep government relationships, could outmaneuver Anduril in the defense AI space despite having a less 'sexy' product.
The second blind spot is the valuation itself. Both companies are priced for perfection. Neuralink's $8-10 billion valuation with zero revenue is a textbook example of a narrative-driven asset. Anduril's 25-30x P/S multiple assumes a decade of hypergrowth. Any deviation from this perfect path will result in a significant repricing. The market is a harsh auditor, and it has no mercy for companies that fail to meet expectations.
Takeaway: The Next Signal
The MSX Pre-IPO fund is a bet on the future of physical AI. The real signal to watch is not the price of the shares but the rate of regulatory and technological progress. For Neuralink, the next catalyst is the expansion of the CONVOY trial and the submission of the PMA application. For Anduril, the next catalyst is the execution of the $2.5 billion DIU contract and the integration of Callum's manufacturing capabilities.
The question is not whether AI will permeate the human body and the battlefield. It will. The question is which company will survive the transition from POC to production. The ledger will remember the data: the number of patients, the number of deployed units, the number of successful missions. Ignore the narrative; watch the numbers. The next quarterly report from Anduril and the next clinical trial update from Neuralink will tell you more than any fund announcement ever will.
