It feels like 6 months ago, I couldn’t go a few hours without being exposed to some new wild claims from Microsoft or Google, or any of the other companies working on this. Lately nothing comes up in any of my feeds.
The things about quantum computers is they are really, really powerful at a some kinds of things, some of the time. Maybe even most of the time.
Binary computers are okay at one thing, but it does that one thing exactly every time. If it ever makes a mistake then entire thing might crash.
The problem is a quantum computer is so different from normal computers you have to start over from the beginning of the tech stack. Even if you get it working reliability (which is hard to do), how does it help you? You can do some crazy math, but it can’t run an operating system. It doesn’t have a bios. There are no drivers that speak some percentage of up spin vs down spin. There isn’t a standard assembly language. There isn’t a standard anything. It’s like how computers were in the 40s.
I think quantum computers are very interesting and might help in very specific applications, and I mean like custom made math heavy applications that only run on that one machine you built it for kind of specific.
Can that change the world? Maybe. But it will take time. No one wants to talk about it because it’s a hard problem that is still years away from doing much to impact peoples lives. Better to talk about ai slop, that gets clicks because you can see it and touch it right now.
So it won’t help me get more fps in counter strike?
It sounds like it’s at the spot that digital computers were at prior to WWII.
Hopefully it doesn’t take WWIII to make them useful.
Well, they are useful for breaking encryption…
Heeere we goooo…
Quantum computers will always exist as a coprocessor, like a GPU or an NPU. You cannot run an operating system on your GPU either. Your GPU is worse for a lot of general purpose tasks your CPU can do, but it excels at very specific kinds of tasks, and so your CPU delegates those specific tasks to the GPU. That is how quantum computers work in practice, they are never stand-alone computers. They are always attached to a classical computer that delegates tasks to it. They won’t ever replace regular computers. Even if in the distant future they manage to build quantum optical chips that run at room temperature at can be made consumer-affordable, they will just be sold as QPUs which you would install into your regular personal computer if you need one.
Everyone working on quantum computers used their newfound breakthroughs to jump to better parallel timelines
It’s AI isn’t it?
Because it was always bullshit. There was never any real progress on it, so they switched their focus to AI and are now swimming in investment capital. No sense backtracking to older scams, when they’re running such a lucrative new one.
I agree. It is hokum and I’ve been saying it for a decade (I have accumulated 10^6 downvotes for speaking this opinion).
But I think deep down I’m snooping to see if people who follow it more closely than I have lately, have detected if the gig is finally up and nobody wants to talk about it?
e: Couldn’t deal with minor typing errors, fixed
e2: Not that the concepts of QC are complete hokum, more refined to say, in my opinion, there is no realistic possibility to extrapolate today’s minor wins and proof-of-concepts into functional tools in the near future. QM by it’s very nature prevents large-scope QC in it’s current incarnation and imagining. Kinda ironic. The scientific community only recently figured out in 2022 that locality isn’t a thing. I feel spacetime is a great model that needs complete overhaul, and SR is done for. Opinion only but QC is cart first horse later.
There’s two problems with quantum computing:
-
They are fundamentally analog machines. Digital computers have huge error bars, its basically impossible for a transistor to bit flip naturally. A small change in voltage does not change the represented value. This is untrue of quantum machines, where a small fluctuation in the qubit’s underlying state actually causes a change to the represented value.
-
We have already done tons of research into what we can do with quantum computers. Theoretical quantum computer science is a well researched field. Basically the only thing we’ve gotten out of it is Shor’s algorithm. Its honestly just not that useful even if we had high breakthroughs in the engineering.
40+ years in tech and I’ve studied QM and QC. I think QC is bogus.
But what I’m genuinely curious is the meta. Why has all the chatter shut down, and what’s actually happening? Is the realization it’s never going to happen finally real?
What you see in mainstream news is always an exaggeration to hype up the layman. Right now all the hype is still focused on LLMs. The recent quantum breakthroughs that were “big news” a year ago were not as exciting as the news might have made it seem.
Quantum research is still going at a slow but steady pace. We have working quantum computers. However, useful working quantum computers have to compete with massive classical computing clusters, which have a huge head start, even considering the theoretical scaling advantages of quantum computers.
“Quantum supremacy” will come. But it also won’t be that exciting because the problems it’s good for are generally limited to niche scientific research scenarios. Maybe really big data centers might find some use for Grover’s search.
The headline to look out for in regards to quantum computing will be something like “quantum computer discovers new material” (with varying levels of exaggerating language depending on the website). That will mark the start of quantum computers being useful (and potentially profitable).
-
They now use AI to boost stocks. But it has not gone away. Just no financial incentive to hype up the progress.
After the AI bubble pops, news will be easier to find
Most of that news was junk, anyway. They run some useless benchmarks, because those machines are incapable to solve real problems.
I’ve perceived the “news” as more a manipulation of users of various SM for the purpose of competitive posturing, and grabbing that sweet, sweet, government-earmarked money specifically set aside for the tech monopolies.
Microsoft is trapped in the ai bubble and wanted to control the pop so that they came out on top. So they timed their quantum news along with other “ai won’t get agi” news to try and manufacture a pop where they stayed high on the hype economy a la quantum. But it didn’t work and we’re still all trapped in the ai bubble.
It’s still way too early for the hype to sustain itself, no one is near making an actual product. Hype is often generated out of the investment world and all the capital is flooding into AI now so their hype industry is largely focused on that for now.
It is all artificial, isn’t it?
First, Its a super specialized technology that requires a lab and paper writing graduate academics to even have real access for stack building.
Then, what are the problems that are even advantageous to solve with a quantum computer. Most daily life problems people face can be delt with by traditional computation. The stuff that really benefits from things like grovers algorithm are cutting edge simulations requiring math thats still being figured out.
For most people the concept and utility of a computer begins and ends with a device to watch YouTube, search internet, play games, send email, do work. Quantum computers are almost explicitly the domain of advanced stem engineers and applied mathematics, things that go over most peoples heads.
Anyone interested in mathematical modeling of hyperdimensional tensor matrixes or to study quantum error correcting codes for funsies likely have institutions backing, education, and connections.
You can buy quantum as pretty much public compute these days. But the price is, well, unaffordable (think 100k USD ++ for a short run) and the wait time for your job to get scheduled is up to 14 days.
Multiple governments dumped billions of taxes into r&d, probably half on fomo and half on bad projections. But nobody has a solution to the decoherence increasing when you add enough q-bits to do something novel classical computers cannot do like Shor’s algorithm or equivalent. They have error correction to try to limit decoherence’s effect on the system, but it requires more q-bits so the fix also exacerbates the problem.
Because someone figured out the last piece of the puzzle, didn’t tell anyone, and now we all live in a post quantum world without knowing it. Reality is now subjective in the deepest sense, and the first past the post is still figuring out the controls. This “time line” everybody hates will either resonate and destroy itself, or coalesce into a peaceful long tail of “just enough”
The simulation folks almost had it, but they were to soon, and looking in the wrong direction
Brought to you by Hideaki Anno
