Quantum Computing and the Future of Technology || Peter Zeihan

We've all been hearing sci-fi tales of quantum computing for decades now, but what will its impact actually look like and how soon can we expect it?

Where to find more?

Where to find me on Social Media?

#quantum #quantummechanics #quantumcomputer

[humanperson]

2:30 - Not quite how a quantum computer works.
A quantum computer doesn't measure an electrons position, or speed within an atom. A quantum computer doesn't do anything with atoms.
A quantum computer measures the spin of a quantum particle. We call these quantum bits: 'qubits'.

3:02 - A qubit is determined by its spin: 'spin-up' or 'spin-down'. A qubit can be in a superposition with a probability of having a certain spin. The information of a qubit is stored in the probability of being 'spin-up' or 'spin-down'. A single qubit can have any probability between 0%-100% of being spin-up or spin-down.
The power of a qubit is that they can also be quantum entangled. This means that the probabilities of 2 qubits being spin-up or down, can depend on each other. This results in processing power that can increase exponentially with qubits, instead of linearly with bits.

3:35 - No! observing the data fundamentally changes the data. It is not still there! We call this wave collapse. Doing this collapses the superposition qubit into regular bits.
The problem is keeping the qubits stable (from collapsing) long enough to measure them. The method we use to do this now is to cool them to absolute 0. These massive freezers are why quantum computers are so large. The advancements in technology that need to happen is to find a way to keep the qubits stable at higher temperatures.


Edit Corrections:
- There are multiple ways qubits can be stored, not just the spin of a particle. These include the shell of an electron inside an atom.
- Not all quantum computers need to be supercooled. However, they have their own issues.

[sangloth]

Peter may be well meant here, but what he said is misleading to just flat out wrong. Others in this thread has pointed out some errors, but in my mind the big one is that Quantum computing can only solve very specific mathematical equations quickly using highly specific algorithms. If those mathematical equations aren't being used, the quantum computer is no better than a general computer (and virtually certainly much much worse.)

Shor's Algorithm - Used to solve current methods of encryption. This will likely be a bombshell once we are capable of entangling enough qubits. I suspect most major power's intelligence agencies are collecting and storing encrypted communication, waiting for the day when they'll be able to read it.

Grover's Algorithm - Used for faster database lookups.

HHL Algorithm - Used to simulate chemistry and solid state physics.

Quantum Fourier transform - It's complicated. A simplified description would be it's useful for analyzing real world data and turning them into equations.

[simon.voggeneder]

I don't understand how quantum computing works but I believe the fundamental misunderstanding concerning quantum computing is that it is the successor to our current silicon transistor computers, i.e. that some day in the future your smartphone will run on a quantum chip. This is not the case. Quantum computing has very specific strengths and the computers require infrastructure that precludes them from being used in a truly portable application. My guess is that quantum computing will be used in specific backend data processing where it shines and all user-faced devices will continue to run on silicon. It's a bit like the once-imagined future of a nuclear reactor in every small thing - it turned out that batteries are better suited to the task of local power supply and nuclear reactors are best used as centralized nodes to generate power.

[Takyodor]

Lot's of low quality/incorrect information in this video. A quantum computer does NOT operate by computing faster, or storing more information. They are only faster than regular computers at certain specific tasks, where you can sort of skip calculations and instead let physics work for you. It's not a problem of understanding quantum mechanics, but an engineering problem to make them practical.

[ogukuo]

It will be good to remember that Peter has never been shy to comment on anything and everything, even if he only has a hazy grasp of a subject.

[MacPoop]

The short answer is: No, quantum computing is not the future of tech. But it IS the future of venture capital buzzwords

[Meton]

A few things you got wrong or missed,

1) quantum computers don't do the same type of calculations that transistor computers do. Quantum will never replace transistors, because everything developed so far (both code and methodologies, ) is based on transistor logic, and your simple fundamental computer instructions like add, subtract, bitwise operations, and many others are MANY MANY MANY times more efficient using transistors than quantum operations. If you're multiplying two numbers, transistors are better. If you're factoring a really huge number into its prime factors, quantum is better. We may reach a point where quantum is cheap enough that you have a "super" computer that has both transistor logic and quantum logic, you will never have a pure quantum computer.

2) Transistors get (have gotten, ) exponentially cheaper with greater scale, both in terms of manufacturing output and in terms of density. It is the exact opposite with quantum. The more quantum bits you entangle, the more complex it gets, exponentially. It is entirely possible that quantum will never have any meaningful impact beyond state level espionage, just due to the capabilities of quantum for cryptanalysis, because of how expensive it is.

3) transistors will never get to the single atom level, because you still need to dope the silicon with (traditionally) boron or phosphorous atoms to make it either P+ type or N- type transistors. The silicon is just the framework to hold the malleable extra electron (phosphorous) or the electron hole (boron).

There is no existing quantum computer that has done any real world calculation faster than silicon transistor computers, and has been many (MANY) times more expensive.

Quantum is not a feasible replacement for the end of transistor shrinking. We are reaching that fundamental barrier, and there are no replacements, even in theory, at this time. There are no quantum computers that can compete with traditional desktop computers, much less "super computers".

[bitbucketcynic]

Processors hit the limit of heat management years ago. Clock speeds never went over 5 GHz because the chips just get too hot, so ever since then they've just been adding more physical cores. They can't go faster so they just go wider.

[stephenderry]

To the guys building the quantum supercomputer, let me save you some time.

It's 42.

[mehdiAbderezai]

You did great explaining the traditional computation, dropped the ball on the quantum.

[winjr]

The CEO of ASML recently said in an interview, that there is a good chance that we will reach a point where the cost of newer process nodes will become unfeasible, sooner than we reach the physical limits of lithography.

[fergieferg]

Good way to know Peter is full of it is that he answers questions on a wide variety of topics as if he is an expert in the field.

[michael.j.mueller]

There is very little correct information in this video. Better leave comments on (quantum) computing to the pros.

[quantx]

peter, this is one subject you did not explain well.. In fact, parts were incorrect.

[Jumper]

It is about certain types of math, like probabilities. Quantum doesn't really do it any better than a traditional transistor computer, just does certain types of math faster.

Ask a transistor computer what 2+2 is and it will immediately come back with a solid 4. Ask a quantum computer what 2+2 is and an hour later it will tell you the answer is approximately 4+/- a small uncertainty range.

But ask a transistor computer where a rubber ball thrown in a room full of furniture will eventually land considering these 100 variables and it will take years to figure out which side of the room is more likely +/- a modest uncertainty.

Ask the rubber ball question to a quantum computer and it will take an hour to give you the answer of 90% likely to end up under the table in the SW corner +/- a small uncertainty range.


The real challenge is getting a hybrid system where the transistors computer interface with the humans understands which problems are out of its efficiency range and tosses the occasional math over to the quantum computer to co-process for it to give the humans a fast an efficient answer with the best odds of being correct.

[AdvantestInc]

The transition from semiconductors to quantum is a journey worth exploring. Keep up the great work in shedding light on this cutting-edge topic!

[PeterLindstrom-xw]

Peter is confusing transistor number with flip-flop number. Flip-flops are the fundamental blocks of memory that can be in one of two states and be made to change states via a signal. They are composed of multiple logic gates, which are in turn composed of multiple transistors. On other words, # transisters != # bits. Tbh, as soon as he said he was going to explain quantum computing to us I was ready for mistakes. For once I'd like him to read a viewer question, say he can't answer that cause it's not his field, and then sign off. As for the rest of this video, there were too many mistakes for me to feel like correcting them all. Except for this: please learn to say "nucleus" correctly.

[johndennis]

Hi. I'm a massive fan of your videos, but you're way off here. You're right that a silicon device relies information taking the form of a string of 1s and 0s. But, you know which bits are 1s and which are 0s (because you made it that way with the program)) - and you operate on them accordingly in your computation. However, in a quantum computer each q-bit is maybe a 1 or maybe 0 with some probability, but you don't know (and you cannot know - unless you look), and so you operate with your program on those uncertain q-bit, after which the answer is all possible answers to the problem with a probability attached to each answer. The most probable answer is the most probably right - and so on. That "when you look at it you change is" thing, isn't a problem - it's how it works (so don't look until its finished) .

[yeroca]

No no no.. Peter is smart, but he totally misunderstands quantum computers.

[NeuroScientician]

I genuinely hope you know more about economy than what you know about tech.