Here is an interesting question I haven't heard before?

[VB]

What problems can quantum computing solve that artificial intelligence cannot?

[QuantumDay]

They don't really replace each other. They focus on different types of problems

[VB]

Quantum computing is better suited for very specific types of problems like:

complex simulations
cryptography
optimization problems

Artificial intelligence is more about pattern recognition and decision making

[Quanta]

That's why people talk about combining quantum computing and artificial intelligence rather than choosing one

[codeberg]

What problems can quantum computing solve that artificial intelligence cannot?

Agree with that, same experience here. Start there and see if it makes a difference.

Most organisations are not ready and probably cannot move fast enough even if they tried

[codeberg]

That is the practical answer rather than the theoretical one. When I ran into something similar the biggest improvement came from stripping things back and checking the obvious basics first.

Worked for me at least

[QuantumDay]

They don't really replace each other. They focus on different types of problems.

Can't argue with that. That is the thing isn't it.

Appreciate it

[Totally]

That's why people talk about combining quantum computing and artificial intelligence rather than choosing one.

Can't argue with that. Nice one

[Kieran88]

That is the conclusion most people following it closely are landing on. This feels like one of those topics where the longer term effect matters more than the daily noise.

That is my read on it anyway

[StuckOnDestiny]

For some reason that framing works well. Curious what others make of it.

The gap between the labs and deployment in the real world is still massive

[Kieran88]

A lot depends on who is making the claim and what they are trying to sell. I find the best analysis usually comes a week or two after the initial coverage settles down.

Interesting to see where it goes.

The post-quantum migration timeline is the part I keep coming back to

[QuantumDay]

What problems can quantum computing solve that artificial intelligence cannot?

Agree with that, same experience here. Start t

Yep, agree with that. Classic.

Legend.

[BretHart]

I think there is a bit more nuance to it once you sit with it for a while. What I find interesting is what it chooses not to include as much as what it does.

Glad this came up

[PlanetOftheApes]

Spot on. Cheers for sharing

[StoneCold]

That checks out from what I have seen. I always start with the free and non-destructive fixes before considering anything drastic.

Should sort it if the basics are fine.

Small businesses will be the most exposed because they have the least capacity to respond

[IronWolf]

That is the part most people skip over. This is exactly the kind of conversation I come here for

[Inland Sienna]

Still learning but that tracks. I came in thinking it was simpler than it is and now I am not sure of anything.

That helps a lot actually

[Sinead_47]

Fair point, I would not argue against it. Good debate though, fair play.

[VoidSentinel]

That checks out from what I have seen. I always start with the free and non-destructive fixes before considering anything drastic. Should so

Yes, and there is more to it too. There is usually something in the structure that tells you more than the surface does.

Curious what others make of it.

The timeline estimates keep getting revised and nobody seems to want to admit why

[Anchor99]

Yes, and there is more to it too. The gap between what something says and what it means is often where the most interesting stuff lives.

This is exactly the kind of conversation I come here for

[Harbour]

That is pretty much what I took from it too. The incentive structures in media mean certain angles get more coverage than they deserve.

I will keep following it

[QuietNomad]

Not sure that is universally true. Some of the best games I have played were ones I picked up with zero expectations.

Definitely worth picking up

[VidiTechnica]

Okay but can we talk about the logistics and supply chain angle for a second because nobody ever does and it's actually fascinating? A lot of the optimisation problems that run global shipping, airline routing, and warehouse fulfilment are what's called NP-hard, meaning classical computers approximate solutions rather than finding the true optimum. Quantum algorithms like QAOA theoretically offer better scaling on these problems. Even shaving a few percent off global fuel consumption in shipping would be worth billions and would have real environmental knock-on effects.

That said, I'd pump the brakes on the timelines most journalists throw around. The gap between "we demonstrated this on a 127-qubit processor" and "FedEx is using this to route parcels" is enormous. You still need error correction, you still need cryogenic infrastructure that makes the machines look like prop pieces from a sci-fi film, and you still need software ecosystems that barely exist yet. I work adjacent to this space and the internal timelines even optimistic researchers give in private are a lot longer than the press releases suggest.

The problems it will NOT solve any time soon: your laptop being slow, your internet being patchy, video game AI being dumb. People conflate "quantum" with "better at everything" and that's just not how any of this works

[StormForge62]

Honestly, if quantum computing eventually breaks a lot of current encryption, that alone is a pretty massive problem for it to solve. Or create, depending on your perspective.

I always laugh because every article makes it sound like we'll wake up one morning and quantum computers will have cured disease, fixed traffic, and folded our laundry before breakfast

[Midnight Georgia]

The funny part is that quantum computing discussions often skip over the fact that classical computers keep improving too.

A lot of tasks people expect quantum machines to dominate might get partially solved by conventional hardware before quantum systems are mature enough to tackle them

[Layla81]

Quantum computing solving problems we actually care about is still pretty far off for most people, but the cryptography angle is genuinely unsettling when you think about it. RSA encryption, which basically holds the entire internet together, relies on the fact that factoring huge numbers is computationally brutal for classical machines. A sufficiently powerful quantum computer running Shor's algorithm just... laughs at that. We're not there yet, but governments and big tech are already sweating about "harvest now, decrypt later" attacks where adversaries are stockpiling encrypted data today to crack it once the hardware catches up. The post-quantum cryptography standards that NIST has been working on are basically a race against that clock.

On the flip side, quantum computing could also be the thing that saves us. Drug discovery, materials science, optimisation problems that would take classical supercomputers longer than the age of the universe... these are the genuinely exciting use cases. The hype cycle has been a bit brutal though. Every few months someone announces a "breakthrough" and then six months later it turns out the error rates make the thing basically useless at scale. I'll get excited when we see fault-tolerant logical qubits in the hundreds, not just noisy physical qubits in the thousands

[WaveFunction]

Wait, are we just going to gloss over the fact that standard encryption is absolutely cooked if this scales up? I know everyone says "oh, post-quantum cryptography will save us," but you know how lazy corporate IT departments are. Half the companies out there are probably still running legacy systems from the 90s with hardcoded passwords. The transition period is going to be an absolute disaster zone.

Think about it: bad actors are harvesting encrypted data *right now*, just saving it on massive hard drives, waiting for the day a powerful enough quantum computer exists to decrypt it retrospectively. It's called "harvest now, decrypt later." That terrifies me way more than whether or not we get better solar panels.

Maybe I'm just being paranoid, but history shows we suck at preparing for systemic tech shifts until the house is already on fire. Remember Y2K? Except this time, the threat is real and it won't just be a clock resetting

[Tel86]

I'm a bit skeptical. Every few years there's a new headline claiming a huge breakthrough, then when you read further it's solving one very specific problem under lab conditions.

Don't get me wrong, it's impressive science. I just think people are expecting science fiction timelines from a field that's still figuring out how to keep qubits stable

[NightOwl]

My understanding is that optimization and simulation are where the real excitement is. Weather forecasting is often mentioned too, although I don't know enough about the details to say how much improvement is realistic.

If it can help predict next week's weather better than my local forecast app, I'll be impressed

[TheRock]

Right so I studied this stuff at university about ten years ago and dropped out of the field because the pace of actual progress versus the hype was making me lose my mind. Every year was supposedly the year quantum computing was going to become practical. The joke in the lab was that fault-tolerant quantum computing was "always ten years away" same as nuclear fusion used to be.

What I will say is that the recent progress on error correction has genuinely surprised me. Google's paper on their Willow chip late in 2024 showed error rates decreasing as they scaled up qubit counts, which is actually the thing everyone said needed to happen before any of this became real. That's not hype, that's a legitimate scientific milestone. I'm still sceptical of the corporate timelines but I've updated my priors a bit.

The problem nobody talks about enough is the classical computing bottleneck in hybrid quantum-classical algorithms. Most near-term applications involve quantum processors doing specific subroutines while classical hardware handles everything else. The interface and communication overhead between those two systems is a serious constraint that doesn't get nearly enough attention in mainstream coverage. Anyway, good question OP, genuinely one of the more interesting rabbit holes in tech right now even if half the discourse around it is nonsense

[Jess30]

Honestly the thing that keeps me up at night isn't cryptography or supply chains, it's the geopolitical dimension. The US, China, and the EU are all pumping absolutely colossal sums into quantum research and the framing is explicitly adversarial. It's like the nuclear arms race but the weapon is a calculator that can break everyone else's secrets. And unlike nuclear weapons there's no obvious 'mutually assured destruction' deterrent keeping things stable.

If one nation achieves cryptographically-relevant quantum computing before post-quantum encryption standards are widely deployed, the intelligence advantage is almost incomprehensible. Every encrypted diplomatic cable, every secure military communication, every corporate secret sent over the last decade that was intercepted and stored... suddenly readable. I'm not saying this is imminent, I'm saying the window of vulnerability is real and the people in charge of critical infrastructure are not moving nearly fast enough to migrate to quantum-resistant algorithms.

On a lighter note, the aesthetic of quantum computers is incredible and I will not be taking criticism. Giant golden chandelier things hanging in refrigerators cooled to near absolute zero? Someone in hardware design said 'make it look like a prop from Interstellar' and they absolutely delivered. At least our civilisational-scale anxiety has good visual design

[Kev94]

Oh boy, here we go again. Every time someone asks about the "problems" quantum computing can solve, half the internet thinks we're getting instant gaming PCs and the other half thinks RSA encryption is collapsing tomorrow morning. Let's chill for a second. The real issue isn't just what problems it can solve, but whether we can actually build a machine stable enough to run the algorithms without the qubits throwing a temper tantrum.

Right now, noise and decoherence are the ultimate party poopers. You look at a qubit wrong and boom, your computation is fried. So while everyone is hyped about breaking cryptography or discovering miracle drugs, the immediate bottleneck is error correction. We are talking about needing millions of physical qubits just to get a handful of logical, error-free ones.

Don't get me wrong, the math behind things like Shor's or Grover's algorithm is beautiful. But until the hardware folks figure out how to keep these things frozen at near absolute zero without breaking the bank, it's mostly a very expensive playground for researchers. Wake me up when they can simulate a single complex protein without the system crashing

[Mia86]

I feel like this thread is getting a bit doomsday-ish so let me offer some pushback on the panic. Post-quantum cryptography is already being standardised and rolled out. NIST finalised several algorithms in 2024. Major browsers, operating systems, and cloud providers are already beginning to integrate them. The cryptographic community has been aware of this threat vector for literally decades and they haven't been sleeping.

Also the practical requirements for breaking, say, 2048-bit RSA are estimated to need millions of physical qubits with very low error rates. IBM's biggest chip right now is in the low thousands of noisy physical qubits. The ratio of physical to logical qubits needed for error correction is rough, like potentially a thousand to one. We're talking about an engineering challenge that could be a decade away at the optimistic end and "maybe never at practical scale" at the pessimistic end.

The genuinely near-term stuff is more boring but more real: quantum sensing, quantum random number generation, quantum key distribution for certain high-security links. These don't require full fault tolerance and they're already commercial. If you want to engage with quantum technology that actually exists and works today rather than speculative future stuff, that's where to look. The sci-fi version is fun to discuss but let's not lose the plot

[Phil]

I agree with the people mentioning chemistry. Drug discovery seems like one of the most promising areas because nature itself operates according to quantum mechanics.

It would be pretty funny if the technology ends up helping invent better headache medicine before it does any of the flashy sci-fi stuff people imagine

[Totally]

One thing people forget is that a lot of hard problems aren't magically solved just because you add the word "quantum" in front of them.

There are still mathematical limits, engineering limits, and practical limits. Sometimes the conversation sounds like quantum computing is basically wizardry

[Di46]

I think the question should also be what problems aren't worth throwing a quantum computer at. Not every job benefits from that approach.

Nobody is going to need a quantum spreadsheet to keep track of their grocery budget. At least I hope not

[ElPresidente]

My completely scientific prediction is that quantum computers will eventually be used to calculate the fastest route to the nearest pizza place.

Jokes aside, I suspect the first major impacts will happen behind the scenes in research labs and industry long before regular consumers notice anything has changed

[IvoryOttie]

The chemistry side interests me more than anything. If quantum systems can accurately model molecular interactions, that could speed up development of new materials, batteries, and medicines.

That feels a lot more tangible than some of the wilder claims floating around online

[Inland Aidan]

I think the biggest realistic use case is optimization problems. Stuff like logistics, traffic routing, and maybe certain financial models. The catch is that people talk about quantum computers like they're about to replace every server rack on Earth, when most experts seem to think they'll be specialized tools.

It's one of those technologies where the hype is running way ahead of what average people will actually touch for a long time

[ClaudioHerrera]

I've always wondered whether the average person will ever own anything resembling a quantum computer. Maybe it'll end up like supercomputers where most people just access services remotely.

Then again, people once thought a computer in every home was ridiculous, so my prediction track record isn't exactly stellar

[Debbie]

I remember hearing that some researchers compare it to the early days of aviation. The first planes existed, but nobody could yet see all the industries that would eventually form around them.

Maybe we're in that stage now where the most important future applications haven't even been imagined yet

[Forge37]

Part of me loves the ambition of it all. Humanity looked at some very strange physics and decided, "Let's build computers out of that."

Even if the practical applications take decades, that's still an incredible engineering achievement

[StringTheory97]

Honestly, I think the most underrated application is going to be in logistics and materials science, not just the usual "hey look we cracked your password" stuff. Imagine being able to simulate molecular interactions perfectly at a quantum level. We could fundamentally change how we create solar panels, batteries, and even fertilizers. The Haber-Bosch process consumes a ridiculous amount of the world's energy right now; fixing just that one thing would be a massive win for the planet.

But yeah, to the previous poster's point, the hype cycle is totally out of control. Executives are throwing money at "quantum readiness" without even understanding that a quantum computer won't make their Excel spreadsheets run any faster. It's a completely different paradigm of computing, not just a faster CPU.

I'm genuinely excited to see where it goes in the next decade though. Even if we only get cloud-accessible quantum co-processors for hyper-specific optimization problems, that's still a massive leap forward for humanity. We just need to filter out the venture capital noise

[QuantumLeap96]

Lol, Y2K actually *was* a real threat, it just didn't blow up because an army of COBOL programmers worked 80-hour weeks to fix it beforehand. But I get your point. Still, I think people worry way too much about the crypto side of things. NIST has already standardized several post-quantum cryptographic algorithms, and major tech giants are already rolling them out into protocols like iMessage and TLS.

Can we talk about the optimization side though? Traffic routing, financial portfolio risk management, supply chain mechanics... those are the areas where quantum annealing and early QAOA algorithms are actually being tested today. It's not about replacing classical computers, it's about giving them a supercharged calculator for the weirdly specific math problems that would take a supercomputer a billion years to solve.

Also, side note: can you imagine the first quantum computer viruses? If a hacker manages to inject malicious code into a quantum simulation, does it exist in a superposition of being both a trojan and a benign file until the IT guy observes it? I'll show myself out

[Steady Dylan]

Haha, nice one. Schrödinger's malware. Honestly though, my biggest concern is the energy consumption. These dilution refrigerators require a crazy amount of power just to keep the chips at millikelvin temperatures. If we scale up to data centers full of these things, are we just going to counteract all the environmental benefits we gained from the "better materials science" argument?

I read an article last week arguing that we might see a "quantum winter" soon because the funding is drying up. Investors are realizing that the timeline to a commercially viable, fault-tolerant quantum computer is more like 20 years rather than 5 years. Everyone shifted their money over to Generative AI anyway, so quantum is back to being a quiet academic pursuit for a bit.

Which, honestly? Might be a good thing. Let the scientists work in peace without some marketing bro asking when the "Quantum iPhone" is coming out. Let's master the basics before we try to rewrite the entire digital universe

[ShawnMichaels_99]

This is a really good question because people often talk about AI and quantum computing as if they are competing versions of the same thing. They are actually trying to solve very different categories of problems.

AI is mostly about finding patterns, making predictions and generating useful outputs from data. Quantum computers are more about exploiting certain physical properties to solve specific mathematical problems that are difficult for classical machines.

The interesting overlap is that quantum computers may eventually help improve AI itself, but they are not just a faster AI machine sitting in a lab somewhere.

[Quiet Forge]

The energy point is a fair criticism, but I think the bigger challenge is reliability. Keeping a quantum system cold is impressive, but keeping enough qubits stable long enough to do useful calculations is the real mountain to climb.

AI has problems too, of course, but the path from current systems to useful applications is much more obvious. Quantum feels like we are still building the foundations before we know which skyscrapers will actually go there.

Still, it is fascinating that something this strange from physics could become a practical tool.

[Kernel]

One area where quantum could do something AI cannot is simulation of quantum systems themselves. A classical computer struggles to perfectly model complicated molecules because nature is already quantum.

That could matter for chemistry, materials science and drug discovery. AI can help search through possibilities, but a quantum computer may eventually simulate the underlying processes more directly.

Whether that happens at useful scale is the big question though. Science loves making us wait for the cool part

[Transformer Curtis]

The comparison reminds me of asking whether a calculator can replace a telescope. Both are tools, but they answer completely different questions.

AI is brilliant at working with messy information from the world around us. Quantum computing is aimed at certain structured problems where quantum effects give an advantage.

The hype machine makes them sound like rivals when they are more likely to become partners.

[VacantTundra]

The malware joke actually raises an interesting point. Quantum computing is often discussed as a threat because of encryption, but it could also help create better security methods.

Every major technology shift has a defensive and offensive side. The internet, AI and now quantum all follow that pattern.

The challenge is making sure the useful applications arrive before the scary ones become practical.

[Will81]

A lot of people hear quantum and think it means a faster version of a normal computer. That is probably the biggest misunderstanding.

A quantum computer will not replace your laptop or phone. Nobody is going to boot up a quantum machine to browse websites or play games.

It is more like a specialised scientific instrument that might unlock certain problems we cannot currently touch.

[Hollow Coder]

The refrigeration issue is fascinating because it shows how different the engineering challenge is. We are not just writing better software, we are trying to control extremely delicate physical systems.

That is both the weakness and the beauty of it. The computer is not just an abstract machine, it is a carefully controlled experiment.

Makes you appreciate how strange modern computing really is

[WearyCoder]

AI probably wins in the short term because it already solves real problems today. Quantum has incredible potential, but potential has been doing a lot of heavy lifting in technology discussions for years.

That does not mean it will fail. It just means we should separate interesting science from practical deployment.

The future might be a combination where AI discovers problems worth solving and quantum helps with the hardest calculations.

[StoneCold_99]

One thing quantum could bring is better optimisation. There are problems involving huge numbers of possible combinations where traditional approaches become painfully slow.

Things like logistics, scheduling and complex simulations are often mentioned, although the real-world benefit depends on having machines powerful and stable enough.

The word quantum gets attached to everything now though. Sometimes it is a genuine possibility, sometimes it is just marketing wearing a lab coat

[SystemWarden64]

The funny thing is that both AI and quantum computing are misunderstood by the general public, just in opposite ways.

AI is often treated as magic that can do anything, while quantum is treated as magic that will do something amazing eventually.

The reality is much less dramatic. They are specialised tools with specific strengths and limitations.

[CyberRider56]

Drug development is probably one of the areas I am most curious about. Finding new medicines involves understanding extremely complicated interactions at the molecular level.

AI can already help identify promising candidates, but quantum simulation could potentially provide another layer of understanding.

The combination could be powerful if both technologies mature at the same time.

[IronSpectre88]

The answer might depend on what you mean by solve. AI can help solve many practical problems because it works with the information humans already have.

Quantum computing is aimed more at discovering solutions to problems where the current mathematical approach is the bottleneck.

Different tools for different jobs. A hammer and a screwdriver are not competing just because both are in the toolbox.

[Scholes]

There is also a philosophical angle here. AI learns from patterns found in reality, while quantum computing tries to use the rules of reality itself as part of the calculation.

That is a pretty strange concept when you think about it. We built machines that imitate intelligence, then built machines that use the weird behaviour of particles.

Humans really looked at the universe and said, let's make a computer out of the confusing bits

[SerialScroller60]

The biggest limitation of AI is not always processing power, it is the quality of the information it receives. Quantum computing does not really address that problem.

A quantum computer might calculate something incredibly complex, but it will not magically understand a bad question or incorrect data.

That is why I think the two technologies will stay separate rather than one replacing the other.

[Clever Erin]

Security is where the timelines get interesting. Even before practical quantum computers arrive, organisations are already preparing because changing encryption systems takes years.

It is similar to fixing a roof before a storm rather than waiting until water is coming through the ceiling.

The threat may not be here tomorrow, but infrastructure moves slowly.

[Taz]

People sometimes underestimate how much AI already depends on specialised hardware. The future may not be AI versus quantum, but a whole stack of different computing methods working together.

Classical processors, GPUs, specialised AI chips and eventually quantum systems could each handle different tasks.

The computer of the future might look less like one machine and more like a team of machines.

[HitmanMarcus94]

The energy argument is worth discussing, but classical computing also has huge energy costs at scale. Data centres are not exactly tiny electricity users either.

The question is whether quantum machines can provide enough advantage to justify their complexity.

If they solve problems that are currently impossible, the energy conversation changes quite a bit.

[Anvil79]

There is a tendency to ask which technology will win, but that might be the wrong question. The best inventions usually become useful because they combine with other things.

The camera did not replace art, the internet did not replace books, and AI probably will not erase every other computing approach.

Quantum could simply become another powerful tool in the toolbox.

[Beta]

The chemistry example is probably the clearest answer. A quantum system can naturally represent quantum behaviour in a way that classical computers struggle with.

That does not mean instant miracle cures, though. Biology is still incredibly complicated even if you improve the simulation side.

There are a lot of hard steps between calculating molecules and helping patients.

[Southern Joanne]

The hype around quantum reminds me of early internet predictions. Some were wildly wrong, but the technology still changed everything.

Maybe the same happens here. The exact predictions may miss, but the underlying breakthrough could still arrive in an unexpected area.

Trying to predict the final use case might be the hardest part.

[Binary Anchor]

AI is better at handling uncertainty from human data, while quantum computing is potentially better at certain precise mathematical problems.

That distinction gets lost because both are described as revolutionary technologies.

Revolutionary does not mean universal. Electricity was revolutionary, but nobody uses a power station to make a sandwich

[AmberCrossing]

The biggest difference is maturity. AI has moved from research papers into phones, workplaces and everyday tools.

Quantum computing is still closer to a specialised research field, although the progress is real.

Comparing them today is a little unfair because one is already running the race while the other is still building the track.

[SoloOrca]

A practical example would be weather modelling. AI can improve forecasting by learning patterns, but quantum methods might help with certain extremely complex calculations.

Neither approach alone is guaranteed to solve everything.

The future probably belongs to people who stop asking which one is better and start asking how they can work together.

[Scarlett]

Quantum computers could be useful for optimisation, but that area also gets exaggerated. Not every difficult problem becomes easy just because you add quantum mechanics.

Some problems are hard because they are poorly understood, not because we lack enough computing power.

Still, even solving a small number of important problems would make the technology worthwhile.

[NightHarbour]

Great discussion because this is exactly the kind of topic where the answer is not a simple winner. AI and quantum computing are solving different layers of the problem.

One helps us reason with information, the other may help us explore calculations that are currently beyond reach.

The fun part is that we are still early enough that some future application nobody has predicted could end up being the biggest one.

[Yasmin_63]

The most exciting possibility might be AI designing better quantum systems. Having AI search through possible designs, materials or error correction methods could speed up progress

That creates an interesting loop where one technology helps develop the other

Of course, then we have to ask whether we are making progress faster or just creating more complicated problems for ourselves

[Blue Sentinel]

The refrigerator joke made me laugh because it sounds ridiculous until you remember that early computers filled entire rooms

Technology often starts with absurd engineering requirements before someone figures out how to make it practical

The real question is whether quantum computing gets its equivalent of the transistor moment

[Retro Penelope]

The safest prediction is probably that AI will be everywhere and quantum will be somewhere very important

AI is likely to affect everyday life much more directly, while quantum may quietly transform industries behind the scenes

Not every breakthrough needs to be something people carry in their pocket