3 ms·

I am surprised no one has mentioned quantum information yet. Quantum information is the subfield of physics whose goal is to understand what new information pro

I am surprised no one has mentioned quantum information yet. Quantum information is the subfield of physics whose goal is to understand what new information processing tasks are possible/efficient that are not possible/inefficient in the quantum world. Two such tasks are quantum computing and quantum key distribution. This field was birthed in the early 80s and has seen steady progress since then. Various architectures for quantum computing have been proposed and experimental control over them has steadily increased in each one over the last few years. While we are not there yet the sequence of results shows that we are rapidly approaching the fault-tolerance thresholds after which it will be possible to build quantum computers. Quantum key distribution is a simpler task and has already been achieved commercially. Now we are trying to increase the rates of transfer. We are slowly also relaxing experimental requirements. For instance there are protocols where you don't have to trust that your devices where not tampered with by an adversary. These are exciting times in quantum information. While important theoretical results where found in the 80s and 90s the experimental momentum today far outpaces it.

mjfl 11y ago
Don't forget the new field of superconducting quantum annealers! I have high hopes.
stared 11y ago
Quantum information is indeed a new field. Yet, I wouldn't compare it with 20s or 70s. In the last 20 years there was hardly any paradigm shift in that matter (don't be fooled by poor science reporting/popularizing or scientist overadvertising their results, in search for fame and grants). And for example, when I as attending conferences in quantum information (my PhD field), people were constantly lowering their expectations and making predictions more humble; much more "10 years ago we said it would be in 10 years; now we can say the same". I don't claim that there is no progress. Just that is depressingly slow comparing to the frontier fields of science in engineering. (If it were fast, believe me, I would have stayed in physics.) Of course, future may be different, but who knows... Compare it with e.g. DNA sequencing, where costs went down quite a few orders of magnitude in the last decade. Or image recognition, where in the last decade things though to be extremely hard (because of problems in 70s) like face recognition, image recognition etc - are now standard techniques.
- abdullahkhalids 11y ago
  I myself am getting at PhD in quantum information. I completely agree that the timeline for quantum computing was extremely optimistic 20 years ago. However, I think current estimates are much closer to the truth. We now have a much better understanding of the systems on which we are building QCs. Estimates of fault-tolerance thresholds have become much better. We have 20 years of experimental progress to extrapolate when we expect to hit these thresholds. I agree that progress in the 20s and 70s was much faster. But they were grabbing onto low-hanging fruit. Building even a 'bad' quantum computer is several orders of magnitude more difficult task than building, say, a 486 processor. The degree of experimental control required is much much more. What I see looking back is steady progress towards greater experimental control in multiple systems and very frequent and steady achievements of milestones. What you forget when you compare quantum computing with DNA sequencing or image recognitions is that quantum computing is at the bottom of current theoretical paradigm. When you do DNA sequencing the science of your instruments (eg. centrifuges) is not suspect. In QC everything is suspect; your system, your detection mechanisms, your control systems. If you normalize every field by its fundamental difficulty you will find that quantum information is keeping up with other fields.
  - stared 11y ago
    I would argue that this difficulty is not more fundamental than difficulty of producing light bulbs by Edison. The thing I do not want to normalize w field by its difficulty - then progress would be just related to the effective number of smart people working on it. I was raised in a scientific culture (Central/Eastern Europe), where difficulty was a virtue, not - impact. But that way produces a --lot-- small number (because they are difficult!) of difficult results, with little impact. Likewise, the most progress happens where there are low-hanging fruits.

mjfl 11y ago

Don't forget the new field of superconducting quantum annealers! I have high hopes.

stared 11y ago

Quantum information is indeed a new field. Yet, I wouldn't compare it with 20s or 70s. In the last 20 years there was hardly any paradigm shift in that matter (don't be fooled by poor science reporting/popularizing or scientist overadvertising their results, in search for fame and grants). And for example, when I as attending conferences in quantum information (my PhD field), people were constantly lowering their expectations and making predictions more humble; much more "10 years ago we said it would be in 10 years; now we can say the same". I don't claim that there is no progress. Just that is depressingly slow comparing to the frontier fields of science in engineering. (If it were fast, believe me, I would have stayed in physics.) Of course, future may be different, but who knows... Compare it with e.g. DNA sequencing, where costs went down quite a few orders of magnitude in the last decade. Or image recognition, where in the last decade things though to be extremely hard (because of problems in 70s) like face recognition, image recognition etc - are now standard techniques.

abdullahkhalids 11y ago
I myself am getting at PhD in quantum information. I completely agree that the timeline for quantum computing was extremely optimistic 20 years ago. However, I think current estimates are much closer to the truth. We now have a much better understanding of the systems on which we are building QCs. Estimates of fault-tolerance thresholds have become much better. We have 20 years of experimental progress to extrapolate when we expect to hit these thresholds. I agree that progress in the 20s and 70s was much faster. But they were grabbing onto low-hanging fruit. Building even a 'bad' quantum computer is several orders of magnitude more difficult task than building, say, a 486 processor. The degree of experimental control required is much much more. What I see looking back is steady progress towards greater experimental control in multiple systems and very frequent and steady achievements of milestones. What you forget when you compare quantum computing with DNA sequencing or image recognitions is that quantum computing is at the bottom of current theoretical paradigm. When you do DNA sequencing the science of your instruments (eg. centrifuges) is not suspect. In QC everything is suspect; your system, your detection mechanisms, your control systems. If you normalize every field by its fundamental difficulty you will find that quantum information is keeping up with other fields.
- stared 11y ago
  I would argue that this difficulty is not more fundamental than difficulty of producing light bulbs by Edison. The thing I do not want to normalize w field by its difficulty - then progress would be just related to the effective number of smart people working on it. I was raised in a scientific culture (Central/Eastern Europe), where difficulty was a virtue, not - impact. But that way produces a --lot-- small number (because they are difficult!) of difficult results, with little impact. Likewise, the most progress happens where there are low-hanging fruits.