What is the holy grail invention of all these fields of science?

[Kaytara]

biology: DNA printer.

Copy and paste a protein sequence, even an entire organism. Extinction would be a thing of the past, you could always make another panda. Create entirely synthetic life. Not just stealing millions of random pieces of another genome and breeding until you find one that works.

Trouble is, wouldn't be quite that simple. Organisms aren't just made of protein, after all - it's sugars and lipids and minerals and complex combinations of all of those, with some parts that can be made by the organism and others that need to be taken up for that to happen, and DNA serves as a blueprint for proteins and RNA (which can have regulatory and catalytic roles as well), but certainly not directly for sugars or minerals. And to paraphrase smuchmuch, we already have highly-efficient 3D printers that are specialised to take up materials from their environment, build proteins and other complex molecules, and release waste - they're called "life". ;) Though I agree that being able to input the right materials and just print a panda would be awesome.

The actual process of making proteins from DNA without a cell involved is called in vitro translation, and it exists. However, there's a big obstacle to equating what happens in a test tube with what happens in the cell. Proteins are ultimately just long strings of amino acids folded in super fancy ways, and they don't start out pre-folded. They fold into the right shape as they are being translated (which may yield different results from if they're allowed to fold spontaneously after being finished) and may require specific chemical conditions (like pH) AND specific other proteins (chaperones) to fold right. Simply making a protein directly from DNA wouldn't resolve that.

[danzibr]

For math, I'd say a thinking machine.

[Mando Knight]

If it ever comes, the problem will be cooling, energy tends to become heat, and the temperature of the Earth is already rising.

Barring some kind of crazy giga-project like trans-atmospheric radiator towers, there's no conceivable cooling method that would reduce the global temperature increase caused by using a positive-power fusion reactor: you're just shuffling heat around, not actually removing it from the system.

[Jay R]

1.) Physics
Far too wide a field to have a single clear answer. It would be easy to say, "Unified Field Theory", but we have no idea if that exists. I suspect that the answer is quantum mechanics, but anybody who thinks he knows the answer has far more ego than I do (an impressive trick, that).

2.) Chemistry
The periodic table. Nothing else comes close.

3.) Biology
Again, far too large to have a single answer, unless it's the recognition of genes and DNA.

4.) Computers
Artificial Intelligence. Which I suspect that we will never reach. (Or possibly the computer chip, which we have already reached.)

5.) Medicine
The cure for cancer.

6.) Mathematics
Calculus. Nothing else can compete.

7.) Geology
Until we know what's under the clouds of Jupiter, we have no business guessing what this could be.

8.) Agriculture
The double bladed plough.

[Eldan]

Placing shades at the L1 point of the Earth-Sun system would do the trick, but it'd likely kill a ton of plant life while interfering with solar energy.

That would reduce the amount of heat we get, but the question was radiating away the heat produced by reactions on Earth. Which is probably not feasible, unless we are talking about something stupid, like taking a space elevator and moving hot objects up it.

For reflecting sunlight, cloud seeding is the more interesting idea to me.

[Errata]

That would reduce the amount of heat we get, but the question was radiating away the heat produced by reactions on Earth. Which is probably not feasible, unless we are talking about something stupid, like taking a space elevator and moving hot objects up it.

We're talking a civilization that uses the energy of a small city for every man, woman, and child on the planet. With that level of technology it's completely feasible. A type 1 civilization would be one that has nearly total control of their planet. Space elevators would be trivial long before that point, and moving physical objects would be a completely unnecessarily elaborate Rube Goldberg way of cooling things. That's a very distant future scenario, not just one or two but many technological revolutions beyond where we are. People are anticipating potential concerns of an extreme distant future while not thinking through all the tools that go hand in hand with the level of technology that would have those concerns. If we had the technology to use that kind of energy, we'd also have solutions for it that are completely impossible now but totally practical with the drastically more advanced technology using that much power. They wouldn't be using that much power just for Playstation 150, they'd be doing actual useful things to require so much more energy, and some of those useful things would be tools for dealing with those sorts of problems that aren't available to us today.

Fusion of the limited supply of deuterium from our oceans is NOT going to be the thing that allows us to generate enough energy that this is a concern, AND it will allow us to reduce the fossil fuel usage that actually causes global warming in real life, so this is an absurd tangent to raise based on a discussion of fusion power.

[Mando Knight]

That would reduce the amount of heat we get, but the question was radiating away the heat produced by reactions on Earth. Which is probably not feasible, unless we are talking about something stupid, like taking a space elevator and moving hot objects up it.

For reflecting sunlight, cloud seeding is the more interesting idea to me.

A solar shade at L1 would still also qualify as a crazy giga-project, moreso due to the size needed (thousands of km in diameter) to have any kind of discernible effect.

[Knaight]

2.) Chemistry
The periodic table. Nothing else comes close.

It helps that the periodic table can subtly change over time, with more and more information added to it (molecular weight, then atomic number, then orbital classification, etc.) - the modern periodic table looks very different from the first one. With that said, the discovery of particle wave duality, VSEPR theory, and the general understanding of how orbitals and bonding work given quantum mechanics comes pretty close to the periodic table in utility. Having studied chemistry, I'm really glad I wasn't one of the unfortunates who had to make do without these tools.

[Peelee]

2.) Chemistry
The periodic table. Nothing else comes close.

Having studied chemistry, I'm really glad I wasn't one of the unfortunates who had to make do without these tools.

For those who aren't familiar with Dmitri Mendeleev, he was amazingly badass. He invented the periodic table, and it was so well-done and accurate that when a French chemist claimed to have discovered a new element, Mendeleev told him that he was wrong, because it didn't have the weight and density that Mendeleev predicted it would. And Mendeleev was right. Without even seeing an as-yet unknown element, he knew what it would be like. Also, his mother trekked him across Russia to Moscow just to get him into school, and when Moscow rejected him, they just continued on the St. Petersberg. These people were dedicated.

[Errata]

Assuming we suddenly became a Type 1 civilization, oceanic deuterium could still sustain us for at least a million years.

Fair enough. That's a lot of deuterium, and it's not all that many orders of magnitude before waste heat becomes relevant.

However, the technology to simply shed waste heat into space may not be so out of reach either: http://www.nature.com/articles/ncomms13729

The atmosphere is relatively transparent to certain wavelengths that have been demonstrated to make space an effective heat sink. With many orders of magnitude increases in human energy use, it would become feasible to scale up radiative cooling drastically.

To make that idea even more efficient, we could bypass a lot of the atmosphere by using a conventional heat pump to circulate refrigeration fluids as high as practical, and then emitting it from the stratosphere rather than ground level. It doesn't require a full space elevator to get a lot of benefits. If we got 10 miles above sea level, the air pressure is miniscule compared to sea level, and we'd be skipping the vast majority of atmospheric interference.

Stratospheric tower concept: http://www.extremetech.com/extreme/2...tors-practical

To be clear, I don't think this sort of project is a helpful solution to our current global warming problem, which involves the rate of heat transfer due to greenhouse gases. It's only meaningful in the hypothetical future of dramatically increased human industry in which waste heat from our energy consumption became the new driver of global warming.

[NichG]

Currently the energy usage of civilization is about 0.01% of what the Earth radiates into space (18 TW vs 170000 TW). So if we decrease atmospheric heat retention by 1%, that basically pays for a 2 order of magnitude increase in the energy usage of civilization, at least in terms of total heat budget.

Local heat production can create a lot of problems before that point though.

[Errata]

Currently the energy usage of civilization is about 0.01% of what the Earth radiates into space (18 TW vs 170000 TW). So if we decrease atmospheric heat retention by 1%, that basically pays for a 2 order of magnitude increase in the energy usage of civilization, at least in terms of total heat budget.

Everyone thus far in the discussion realizes that. However, some people are speculating about exponential growth continuing indefinitely, which is really putting the cart before the horse in a discussion of new energy sources that can reduce greenhouse emissions and reduce global warming.

[Errata]

The tower being used for space elevators seems to make it vulnerable to orbital debris, and it also seems that significant punctures would lead to destabilization from the pressurized gas spewing out. They admitted that hydrogen is flammable, and helium is expensive, but ignored the potential risk of a catastrophic collapse of a space elevator, due to the upper end getting disconnected from the counterweight.

There is no counterweight, nor orbital debris. Because it's not a full space elevator, it's just one example of a proposal for a tower to the stratosphere, not to orbit. A catastrophic collapse of a tower can be limited to a relatively narrow footprint, since it's not subject to significant orbital forces like a space elevator. And since this particular one would deflate gradually, like a balloon, it would not crash at terminal velocity for its lightweight materials, although it would probably still not be a great situation if it collapsed.

A tower is a compromise vs a full space elevator, since the radiative cooling would work from sea level, but work even better higher. A full space elevator would work through entirely different technology, but would probably also be feasible before we have the industrial base to consume thousands of times more power than we do now. We would presumably have access to much better technology for mitigating risks, and would choose their placement carefully with those risks in mind.

Cost of materials in 2017 doesn't mean much. If we're assuming massive progress, we'd have to assume more than linear progress in our manufacturing capabilities. Steel and aluminum used to be expensive and made in small batches. Between 1900 and 1950, steel production globally grew by a factor of 5x. Between 1950 and 2000, it grew by another 5x. Energy use isn't the only thing that we might expect to grow exponentially. And if anything manufacturing is more than linear with energy, because with today's technology we could scale it up at a similar ratio between energy and productivity, but as our technology improves we find ways to make better and cheaper materials more efficiently. As long as we aren't relying on materials that require a lot of rare minerals, then we can expect to build bigger and better more efficiently in the future than now. When it comes to new materials like carbon nanotubes, we're getting way better at manufacturing them reliably, in longer and greater quantities every year.

[halfeye]

I can't imagine how we'd successfully mitigate the collapse of a structure that's dozens of kilometers tall. Even then, its height would be around 90% of the Earth's circumference

The atmosphere is tiny. The stratosphere is about one dozen kilometres up. The circumference of the earth is several tens of thousands of kilometres, it's not the same sort of scale at all.

[Lord Torath]

9. Materials Science: A lightweight (SG:1 or so) strong (interlocking Carbon nano-rings strength), self-repairing material feasible for use on micro or macro scales.

I'm really just guessing, here. I'd like a material strong and light enough to serve as a Space Elevator ribbon.

[Leewei]

1) Physics - Unified Field Theory

2) Chemistry - Room temperature superconductors

3) Biology - Creation of single celled life in a test tube from raw ingredients

4) Computers - Artificial Intelligence

5) Medicine - Immortality

6) Mathematics - P=NP

7) Geology - A complete model of the formation of planets

8) Agriculture - Cheap, tasty, artificial meat

[georgie_leech]

Oh, woops. Forgot to add the "thousands" part. Thanks for pointing that out.

I think you're missing the part where we don't need something tens of thousands of kilometers tall to aid in radiative cooling. Scale back your estimates by three orders of magnitude; tens is sufficient on its own.

[wumpus]

One catch with claiming a goal of "AI" is "anything a computer has actually solved is no longer a criteria for AI". No idea if "passing the Turing test" will be retro-defined as "not sufficiently AI".

An obvious "holy grail" of computers is parallelism. Compilers break modern computer programs into streams of individual instructions and typically have a best case of computing two of the instructions every cycle (i.e. 6 billion instructions on a 3GHz machine). Progress in speeding up both the IPC (the number of instructions at once) and the clockspeed have essentially stalled with no expected progress. However, ever since the most powerful processors could fit on a single chip (mid 1980s?) it has been obvious that the cheap means of speed is to simply add more processors, and since about 2010 that has been the *only* real means of significant increases (basically the difference between an i3, i5, and an i7 as well as the various AMD Ryzen chips). To put it bluntly, a $200 Intel chip can be expected to run a program at ~6 billion operations/second and have three times the capacity to run any other programs that might be running as well (it rarely needs that much). Nvidia, on the other hand, builds "computers" for problems that we *do* know how to massively parallelize (mostly computing 3d graphics, and only after the CPU does a lot of specific work to build the scene) problems and operates roughly ~1000 times faster (note: the CPU also can typically speed up dramatically with such "embarrassingly easy" parallelization problems, but still is at least ~50 times slower).

When people first encounter the above problem, it seems like parallelism should be right around the corner. But "build it and they will come" has been tried since the 1980s (look up Thinking Machines Corporation for an early serious attempt) and progress has been glacial.

Other problems (not all at "holy grail" level, but often "Bill Gates level money is on the table for the winner*").
biology/marine engineering: I've heard the "fouling problem" is huge. I'm assuming the issue is that any water that has intakes and exits will produce bacteria growth/slime that will continually build up.
Agriculture/chemical engineering: ethanol production from cellulose stock. There has been a ton of money poured down this hole (and huge markets ready to embrace production) but no economic success.
chemical/battery tech: Flow batteries. Basically "rechargeable fuel cells", these require far less "expensive bits" of the battery and allow arbitrary amounts of chemicals and vastly smaller bits of the rest of the battery. So far, no luck. Note that it isn't completely obvious that they would be always better than other ways to cut the cost of "the expensive bits". Zinc tech might prove these unnecessary.
traditional car engine tech:HCCI (making a gas engine ignite like a diesel), something Mazda plans to introduce in 2019. Variable compression, coming from Infiniti who knows when. In practice, neither tech will add much more than 20% efficiency (and then only in limited functions), so in practice the "holy grail" is likely a cheap, lightweight, and powerful battery (see "flow batteries" as a possibility) plus a cheap, lightweight and powerful electric motor (and use something like a prius engine to keep the battery charged. Obviously adding one or both techs listed above will help your range extender but in practice the more powerful the engine, the worse the mileage (at least for cars with levels of power that sell in the US), but this really doesn't limit electric motors.
cheap multichip modules While chips have unbelievable levels of communication available inside a chip, communicating with their neighbors is a vastly different story. One method that has been tried is simply stacking chips together, and that has worked well for memory chips (until they need to output to a CPU, then problems). The recently released (and basically snatched up by cryptominers at over list price) AMD Vega graphics cards use "high bandwidth memory", which basically attaches stacks of ram to the processor (a GPU in this case) via a larger, primitive chip that both are sitting on. Intel has recently announced that they have a process similar to the above, but they only need the tiny slivers of the 'large primitive chip' that lies underneath the two chips that are communicating. Obviously this is a manufacturing ability of mind-boggling precision (even if it largely relies on self-alignment), but it should be in production soonish. Compare to the old, dead dream of wafer-scale computing (which was presumably dead when memory and logic started using different processes). Note that this is hardly the "holy grail" issue like others, but it does bring to mind the idea of massive clumps of cores, connected directly to large DRAM caches which are in turn backed up by another [stalled] intel "holy grail" tech, 3dxpoint (apparently it has limited endurance. It might replace a lot of flash tech as is, but it won't be replacing (some) DRAM without the endurance and the price isn't likely to come down any time soon to only replace flash).
* not that I'd expect a measurable portion to go to those responsible for the breakthrough.

[Lord Torath]

chemical/battery tech: Flow batteries. Basically "rechargeable fuel cells", these require far less "expensive bits" of the battery and allow arbitrary amounts of chemicals and vastly smaller bits of the rest of the battery. So far, no luck. Note that it isn't completely obvious that they would be always better than other ways to cut the cost of "the expensive bits". Zinc tech might prove these unnecessary.

Progress is being made. Liquid Metal batteries are extremely durable, and can charge and discharge repeatedly for up to 15 years without degradation. At least according to the article in my "Mechanical Engineering" magazine (Sept 2017). But they have to operate at high temperatures, so don't expect this to power your iPhone or laptop any time soon.

Also, people are investigating using renewable power to create hydrogen using electrolysis, and then combining that hydrogen with CO₂ from the air to produce methane, which is much easier to store than hydrogen. Burning it releases that CO₂ back into the air, but it's the same amount that was pulled out earlier in the process, so no net effect on the amount of greenhouse gasses in the atmosphere. Assuming, of course, no major methane leaks.

[Grinner]

The Manhattan Project showed us many things, among them that it's possible to convert a relatively tiny amount of mass into a profusion of uncontrolled energy. What of the possibility of converting a profusion of controlled energy into a relatively tiny amount of mass?

[halfeye]

The Manhattan Project showed us many things, among them that it's possible to convert a relatively tiny amount of mass into a profusion of uncontrolled energy. What of the possibility of converting a profusion of controlled energy into a relatively tiny amount of mass?

The ratio of energy to mass is too big for that to ever be worthwhile, if it was possible. If you mean as a way to get rid of excess energy, then maybe, but the efficiency is likely to be so low that you create more spare energy doing the conversion than you turn into matter.

[georgie_leech]

The Manhattan Project showed us many things, among them that it's possible to convert a relatively tiny amount of mass into a profusion of uncontrolled energy. What of the possibility of converting a profusion of controlled energy into a relatively tiny amount of mass?

That's pretty much what happens in a particle accelerator. We get all sorts of exotic matter out of it, but it tends to not stick around forngery long and the Second Law of Thermodynamics is a harsh mistress.

[gomipile]

Just one for each category.

6.) Mathematics

•A proof or refutation of the Riemann hypothesis.

•A proof of the minimum possible computational complexity class of an algorithm to factor an arbitrary N bit integer.

[Bohandas]

Biology: Reanimation of dead organisms, elimination of aging, from scratch artificial life, and food replication

Physics: Cold fusion and also negative energy density and its various applications (alcubirre drive, wormholes, etc)

Computers: Brain to machine interface, quantum computation, non-decaying data storage