The biggest question in technology investing right now isn't whether AI will keep growing — it's what comes after Nvidia. That framing explains almost everything about why quantum computing stocks stay inflated despite weak fundamentals, why photonic computing deserves serious attention, and why a founder with a track record in biotech and enterprise software is now trying to build a photonic computing company from the ground up.
Why Quantum Won't Die (Even When It Should)
Quantum computing stocks are, by most rigorous measures, ludicrously overvalued. IonQ and its peers trading at $10–20 billion market caps have no credible path to justifying those valuations on any near-term commercial basis. And yet the stocks persist — and for a specific reason.
Nvidia is the largest company in the world by market cap, worth roughly $5 trillion and possibly heading toward $10 trillion. That extraordinary wealth creation has made investors desperate to find the next Nvidia — the next paradigm shift in computing that could compound over decades. Quantum computing sounds like next-generation computing, so capital flows in. When you try to correct the record with facts, investors respond reasonably: you don't know what's going to happen in 10 years. They're not wrong. But that uncertainty cuts both ways.
The more honest framing is that quantum computing faces fundamental physical challenges that no amount of investor enthusiasm resolves. Quantum remains a short thesis. What's harder to short is the narrative of next-generation computing itself — and that narrative won't collapse until Nvidia does.
The Physics Case for Photonic Computing
If quantum is the wrong answer to the post-Nvidia question, what's the right one? Photonic — or all-optical — computing has emerged as a genuinely interesting candidate, and the reasons are grounded in physics rather than hype.
Electrons moving through copper wire travel at roughly five orders of magnitude slower than photons. Resistance creates heat and latency. Photons, by contrast, move at the speed of light regardless of the medium. That's not a minor engineering improvement — it's a fundamentally different substrate for computation.
More importantly, matrix multiplication sits at the core of modern AI workloads. It is the dominant operation powering every large language model and neural network, and it is precisely why Nvidia's GPU architecture became so valuable. Optical computers can perform matrix operations — sometimes called matrix moles — with a different computational complexity class than GPUs. If that advantage holds at scale, the energy and speed implications for AI infrastructure would be enormous.
This isn't a near-term story. The honest timeline is decades, not years. Companies in this space will face capital constraints, personnel challenges, and the ever-present risk of scientific dead ends. But Microsoft's active investment in photonic computing — directly targeting AI workloads — validated the thesis enough to move the needle from curious to genuinely interested.
Building QCLS: A Photonic Computing Venture
The company in question is QCLS, a publicly traded entity being repositioned around photonic computing. It currently holds a licensing agreement with Israeli photonics firm Lightolver as its primary asset — a thin foundation, but a real one. The plan is to use that base, raise capital through the public markets, assemble a research team, and build toward something meaningful in the all-optical computing space.
The appeal of a publicly traded vehicle is straightforward: capital raising is easier. The risk is also clear — small public companies burn cash, face shareholder pressure, and operate in full public view while doing inherently long-horizon scientific work. The involvement here is as a consultant and operational contributor, drawing on experience building and running companies rather than as a pure capital allocator. The comp structure reflects that: there's no meaningful upside unless the company achieves something genuinely large.
Photonic components already exist on modern motherboards in limited form. The vision is all-optical at scale — eliminating the electron-based bottleneck entirely for certain workload classes. Whether QCLS gets there is unknown. Whether the problem is worth working on is not.
Intel: A Value Stock Facing Structural Headwinds
38:45
Goodell financial model showing Intel earnings estimates, revenue forecast, and valuation multiples
Watch at 38:45 →
A quick look at Intel through a financial model surfaces a stock with genuine value characteristics but serious structural problems. The core question is whether Intel can return to earning several dollars per share. The answer depends almost entirely on whether revenue growth resumes.
At $40, Intel trades at roughly 68 times forward earnings — not a steady-state multiple for a business this cyclical. At $20, it was arguably compelling on a value basis. At current prices, the upside is much harder to justify without assuming meaningful growth, and growth is precisely what's in doubt.
The headwinds are well-documented: AMD continues to take CPU market share, Intel has no credible GPU product to participate in the AI infrastructure buildout, and the broader shift of workloads toward accelerated computing reduces the centrality of the CPU. The CEO is capable and the US government's CHIPS Act investment has shored up the balance sheet — net cash is negative $7 billion, which is manageable given the subsidy environment. But a 3–5% revenue growth assumption, which looks modest, could easily flip negative in a downcycle. Wall Street rarely forecasts cyclical downturns in advance, which means consensus estimates for Intel almost certainly overstate near-term performance.
The honest conclusion: Intel is not attractive at $40. It was interesting at $20. The turnaround thesis requires deeper work on their fab strategy and advanced packaging roadmap before forming a high-conviction view.
On Trading, Transparency, and Refocusing
The challenge account that ran from $30,000 to $260,000 and back to zero was an instructive failure. The post-mortem is honest: leverage amplifies both the upside and the terminal risk, and a single bad position — in this case a heavily weighted short in a biotech that had previously misrepresented trial data — was enough to erase the gains. The position size was too large. The research process had been partially delegated rather than fully owned. And the nature of leveraged trading means there is no recovering from a wipeout.
The broader lesson isn't that trading is impossible — it's that streaming trades, managing outside capital psychologically, and running concentrated leveraged positions simultaneously creates compounding cognitive load that degrades all three activities. The content started as stock analysis. The drift toward live trading performance was audience-driven, not strategy-driven, and the results reflect that.
The path forward is cleaner: stock analysis using purpose-built software, a genuine operational role at a photonic computing startup, and biotech idea generation drawing on two decades of domain expertise. The challenge account is closed. The quantum shorts remain. And the photonic computing thesis is, at minimum, worth the next several years of work.
