What's next for quantum?

I get the feeling that quantum computing is a siren call - an enchanting technology that promises the world (cancer cure, anyone?) but ultimately lures sailors (venture capitalists) onto the rocks. I'll always be enthralled by this technology and track it closely, but I’m taking a few steps back, tying myself to the mast and putting my ear plugs in. I explain Why here.

I don't like bubbles

The quantum computing industry is in a proper state of froth right now. Big rounds at big prices have been routine in 2026. Oratomic raised $475M at a $5.4B valuation three months after raising at $1.5B, without shipping a machine. Nord Quantique hit a $1.4B valuation on a $30M round. Depending on who's counting, around 17 quantum companies are now valued at a billion dollars or more, and six quantum companies went public this year alone: five by SPAC, plus Quantinuum's $1.68B IPO at a $15.6B valuation, on about $31M of revenue. For the most part, these billion-dollar (or ten-billion-dollar) companies don't have any real revenue. The biggest of them, IonQ, spends about $5 for every $1 of revenue it brings in.

I've written before about why these valuations are completely detached from reality. It's clearly a bubble and I don't know a single person working in and around quantum who doesn't think so too.

Progress is being made, but it's slow going

You could be forgiven for thinking all this investment reflects a technology that has been de-risked. There has been progress, but it has been incremental and we are still a long way from knowing (1) whether this technology can be built at all, and (2) whether there are clear economically valuable things to do with it.

Look at the last decade:

• Qubit counts: in 2016, IBM put a 5-qubit processor on the cloud. Google’s Sycamore had 53 qubits in 2019, and Willow had 105 in 2024.
• Gate quality: the best two-qubit gates on commercial machines have gone from about 99% to about 99.9% (Quantinuum’s Helios reports 99.92% across 98 qubits). That is one extra “nine” in ten years.
• The companies themselves are not young. D-Wave was founded in 1999, Rigetti in 2013, IonQ in 2015 and PsiQuantum in 2016. These are teenagers and twentysomethings with the metrics of seed-stage start-ups.

The biggest shift has been in error correction -- a process required to make quantum computers work. For example, the estimated number of physical qubits (with 99.9% fidelity) it would take to break RSA-2048 (a standard benchmark for proving a fault-tolerant machine can do something current computers can't) has moved from about 200 million in 2012 (Fowler et al) to about 20 million in 2019 (Gidney and Ekera) to 72,000 this year (Iceberg). This is super cool and significant but we are still 1000x off a useful machine.

Even if we build the machine, it's not clear what we'd run on it. The algorithms side has had limited progress and the best algorithms (Shor's and Grover's) are 30 years old. Breaking encryption is the clearest use-cases, and you can't really build a business out of that (well you could, but it's not why we all got excited about quantum).

So, what's all the money chasing?

1) Compute. Demand for differentiated compute has skyrocketed on the back of AI. That demand doesn't translate directly to quantum (it's not clear quantum would be useful for AI training or inference at all), but it possibly has made investors far more willing to take the risk, especially when "Quantum + AI" press releases are everywhere. Quantum stocks trade like a leveraged bet on the AI trade.

2) Big tech's spare change. Big tech profits and stock prices have surged, so throwing a little at quantum is no bother. The hyperscalers will spend something like $725–828B on capex this year. All quantum startup investment in 2025 was $12.6B. It makes sense that these groups would like to take a bet on the future of compute, but we shouldn't take it as a high-conviction signal.

3) Retail. Frothy markets have let quantum companies sell retail investors on billion-dollar valuations in the public markets, which in turn has returned early-stage investors' money, which has let them launch new venture funds that invest in more quantum.

4) Geopolitics. Governments have put money into quantum both from a position of fear (quantum computers pose a cybersecurity risk) and because they don't want to miss out on the "next semiconductor industry". Geopolitical tensions mean that more blocs are competing against each other rather than working together, resulting in more public money in the system. That has let quantum companies scoop up extra cash, which has encouraged more private investment and lent the sector legitimacy.

All that is to say, the surging investment does not mean quantum is anywhere near as close to being real as the money suggests.

My prediction

Something will pop the bubble, but it may not need a pin. Bubbles often deflate under their own weight if they run out of new buyers, and they create their own supply as every company rushes to list or to raise a mega-round while the window is open. Both are already happening in quantum with ten public pure plays, a stream of share sales, and Quantinuum’s lockup ending on December 1. If an outside trigger arrives, my bet is it comes from AI. Quantum has been trading as a leveraged bet on the compute boom, and if or when that trade cools, quantum will crater.

This may be the starting gun on the final sprint for quantum. With no more money coming into the sector, public quantum companies who have financially engineered the past few years and companies who have completed mega-raises will have a nice cash cushion to go hard a the goal - a real quantum computer by 2033. Those who do not, like early stage startups, will suffer the worst.

I'd rather stay out of it

Quantum still might happen. DARPA's Quantum Benchmarking Initiative is testing whether any approach can reach useful scale by 2033 and are cautiously optimistic, and plenty of company roadmaps say 2030. But it's still a venture bet being propped up as a guaranteed thing. The sector is feeling silly. More and more venture funds are being launched off the back of paper returns and ethically dubious SPACs. More and more people are pretending they're experts. Everything about quantum just feels a bit...icky.

Too much money, too many "experts", too many companies untethered from reality and frankly, there are too many other interesting things to focus on.

When will I come back in?

I’m not plugging my ears forever. I’ll come back when a companies are buying quantum computers because they generate a return on investment, when someone runs a useful algorithm on 100 logical qubits, or when someone manufactures a thousand high-fidelity qubits that can all talk to each other. Until then, I’m staying tied to the mast. The song is beautiful but the rocks (and opportunity costs) are real.