[kibwen]

Ironically, given that the primary use case for developing quantum computers is breaking classical encryption, switching en-masse to post-quantum schemes may very well ensure that developing a working quantum computer may never be economically viable to develop.

[tzs]

A large part of the investments in developing quantum computing are for drug discovery, materials science, and logistics.

[kibwen]

Certainly, we can imagine other use cases for quantum computers. But just because something could be useful doesn't mean it will be developed; the cost still needs to be surmountable.

[tzs]

That wasn't just a list of things they could besides breaking crypto. It was the list of the main reasons people are investing in trying to build them.

As you noted the problems they pose for cryptography can be addressed with PQC. Military/Defense/National Security invests in them for the cryptographic applications (and probably also for the logistics applications), but they are only about 1/3 of the investment.

The rest is for the doing better physics simulations and for logistics and financial applications. The physics simulations will be huge and drug development and materials science.

[nemo]

Logistics companies are using hybrid QC now from D-Wave Systems. Quantum computing is currently in use in the real world. There's already physics simulations running on QC hardware beyond the capacities of classical computing, QuEra's Gemini is already doing materials science work now. It's very expensive, giant logistics companies and other industries with lots of capital and the need to do the special kinds of simulations these things are good at will be renting time on these things from specialized vendors who can run this kind of intensive hardware, something like time sharing on a supercomputer. Quantum-as-a-Service (QaaS) via platforms like Amazon Braket, Azure Quantum, or IBM Quantum Platform is already a thing. Big Pharma will be renting this stuff, Google currently plans to have QC for Pharma applications running c. 2028-2030. Lots of folks dealing with materials sciences and so on will be on projects where this will be affordable since not everything needs a massive CRQC to be useful.

[semi-extrinsic]

Could you post a few links to these things? Last I remember there were a few PoC around pushing toy-scale examples, but nobody close to "quantum supremacy for application X"?

A cursory google still seems to indicate people are using these things to learn how a hypothetical QC could be used in practice, rather than actually doing stuff in production or anything close to it.

[plopilop]

I have seen https://pubsonline.informs.org/doi/10.1287/mnsc.2023.00314 as an example.

> Despite the limitations in size and speed of today’s quantum computers, our algorithm provides quantifiable liquidity savings when applied to the Canadian HVPS using a 30-day sample of transaction data. By reordering batches of 70 payments, we achieve an average of Canadian (C) $240 million in daily liquidity savings, with a settlement delay of approximately 90 seconds

It has to be noted that the technology used (quantum annealing) is at best erm disputed, and that the company DWave has made very wild claims in the past. Also note that many "quantum speedups" have been de-quantumized, i.e. classical algorithms with equal or even better performance have been developed, sometimes by drawing inspiration from the quantum algorithm. Quantum supremacy is still quite unclear.

Now, could the problems we are talking about be efficiently solved on a classical computer? Maybe. But if nobody knows how to do it, we might just as well use the quantum computer.

[NetMageSCW]

You said the primary use case was breaking encryption - you have a cite for that, or these “other use cases” actually why quantum computing is being developed? Certainly, the public position is they are for breaking encryption but that seems woefully uninformed.

[dredmorbius]

The history of diligence in properly utilising and applying security measures strongly suggests that QC will in fact provide positive utility even if viable PQC countermeasures are developed and deployed.

From yesterday's posts: <https://news.ycombinator.com/item?id=48983610> (my comment on the Romanian land registry hacking thread).

[kibwen]

I'm afraid I don't see how the linked comment demonstrates the positive utility of QC.

[dredmorbius]

The link shows that well-established security practices are often not followed, even by those who really should damned well know better.

Which is to say: even when PQC mitigations against QC are available, I expect a target-rich environment for QC attackers.