Hacker News Reader

Comment by nautilius | Hacker News Reader

The simple answer is: funding. There is no (public) money to revisit "known" ideas and results to help clarify them, apply them, or place them in better context. And I would not be the one to explain to the public to spend their money on known ideas rather than discover something new and shiny.

Concerning a PhD, I think it's pretty easy to see that that's not the stage in your career where you would want to embark on long-term, high risk research.

Also, I am not surprised that people doing a similar job have a common vocabulary and vernacular. You could say culture. I'd be surprised if carpenters, mechanics, et al. don't have the same thing, with other words and topics, of course.

No conspiracies needed.

Wheeler had significant contributions, had the right pedigree, the right schools. I don't see how he would be at a disadvantage today.

View on HN

It's not about conspiracy theories. It's about reality. You even explicitly called it out: culture. That's just the way humans work, for better or worse. One of my points, and I think a point of Disciplined Minds, is that scientists are sometimes not honest or self-aware of the culture that they're a part of. Science is not a purely rational endeavor. There's some fashion and belief to it.

> Wheeler had significant contributions, had the right pedigree, the right schools. I don't see how he would be at a disadvantage today.

I didn't mean it as a critique upon Wheeler's potential. He was incredible. But he switched fields and research directions several times during his long and successful career. I'm not sure academics today have that fluidity in their careers (not due to internal forces but rather the external ones).

I don't see why scientists would be less self-aware of their culture than other groups. In particular, that culture has got nothing to do with the things you decry (why is no one looking at old ideas, where, again, the simple problem is money).

As soon as Wheeler would have tenure today, he could switch all day long. I am not talking about Wheeler's potential either, but whether he ticks the boxes to make him likely to succeed today. And he does.

You are looking for a career that wouldn't work today? Freeman Dyson. Getting tenure at Princeton aged 29 without a doctorate, not so likely today.

> I don't see why scientists would be less self-aware of their culture than other groups.

In my opinion, there’s a rationality bias. In that, if one thinks that they are operating rationally, then they think they are somewhat immune to cultural biases and inclinations. If I am not mistaken, this is similar to things Paul Feyerabend discussed. This is also heavily discussed in the book Disciplined Minds I linked above.

> You are looking for a career that wouldn't work today? Freeman Dyson. Getting tenure at Princeton aged 29 without a doctorate, not so likely today.

I absolutely agree. I think it’s beyond not likely. It is basically an impossibility.

Feyerabend may have a point here, but elsewhere on this site I have raised my doubts about his understanding of some physics notions. It is sad to see errors propagated at the same time as good points.

Disciplined Minds is a good book.

>And I would not be the one to explain to the public to spend their money on known ideas rather than discover something new and shiny.

Speaking of relating scientists to labor workers..

There is a lot of value to be earned by closing the gap between cutting edge, but proven, science and easily understood training documents for contract workers.

I would say most people would not care if we stopped using tax dollars to fund string theory research or finding 'habitable' planets, and a host of other luxury topics. we need to fortify our infrastructure (in every sense of the word) with all the things we have established as 'known ideas' today. it doesnt feel like we are in the golden age anymore where we can afford luxury research. at least not this decade.

imo The only 3 research fields still worth investing tax dollars in is long term energy storage, AI, and GMO. Everything else can kick rocks and start pitching in with more necessary labor or do it on their own time. the reason for those exceptions is that based on what we already know, there is a potential for resolving some unknowns that would result in global disruptions. IE self driving cars have proof of concept, resetting cell age has been done in mice,and the energy crisis demands a need for energy storage research to be exhausted

This standpoint is very specific to the world we live in today. Such luxury research is the reason we are where we are today and we should continue it in the future, but given what we know now - we should be in a heavy transition period towards getting the most out of what we already know. We've reached the part of the brainstorm session where the new topics being introduced arent exactly wrong, but it is borderline annoying and distracting in relation to accomplishing anything and it would be nice if our smartest people would contribute to the project

> we should be in a heavy transition period towards getting the most out of what we already know

That's what a free market is for. Developing technology from what we already know is profitable and self sustaining. I can't think of anything worse than governments directing technology development.

Science is the process of uncovering and understanding things about the world that we don't already know. Who exactly is in a position to know what counts as luxury? By your logic the discovery of the electron 125 years ago was luxury research, and the fact that our civilization is based on electricity and electronics today is irrelevant. General relativity didn't have practical applications for almost a hundred years, now we all rely on it indirectly everyday for GPS.

>what counts as luxury

Paying money to shoot in the dark for miracle theories is a luxury. If you dont have enough food, and you are paying to research how to get more food - it's not luxury research. If you are paying to study string theory instead of importing more food because "you never know what we might learn, maybe world hunger will be solved" then I consider that a luxury. If we disagree on that it's okay and we can move on to other points as that is a fringe point.

No one funded general relativity, they funded Einstein. And the government did not fund Einsteins early theoretical papers AFAIK. To use him as an example is disingenuous. He had already made significant and applicable contributions (which is an understatement). He was definitely funded by a government (as part of a team) to create new technologies which in turn won WWII for that government.

I am not aware of any scientific evidence that government funding is necessary to produce groundbreaking theories. In other words, it's possible that genius happens and self-motivates without prior good-faith funding. I do see historical precedent for huge technological advancement through paying theoretical physicists to stop working on theory and start working on application though.

>I can't think of anything worse than governments directing technology development.

Why? Governments historically have funded an incredible amount of successful bleeding edge technologies. Especially in the defense sector. The atom bomb is one of them.

Part of my point here is that we need to take the effectiveness of what our government is capable of with technology in terms of defense and apply that level of competency to other areas - incentivizing the relevant researchers to be hands-on with application development

Good news, energy storage has plenty of private funding, and mainly just needs to get built out, along with solar to charge it from. For serious bulk long-term storage it would be hard to beat ammonia, except that people will keep trying to buy up your stock to use for other things.

What needs work is economical ways to get carbon out of the air into a form where it will stay. I had thought moving surface water saturated with it down deep was a good idea, but apparently there is more down there than at the surface! So, scattering millions of tons of olivine on beaches and cropland might be the best we can do.

Do we really need AI for anything?

>For serious bulk long-term storage it would be hard to beat ammonia

Is this related to flow batteries? Right now I think pumped storage is winning (where applicable), but I've heard flow batteries are more powerful

>What needs work is economical ways to get carbon out of the air

Isn't it kind of wild that too much carbon is a problem? A core building block for life on earth. I agree though, next decade is important for carbon research as well. Olivine is interesting

https://energsustainsoc.biomedcentral.com/articles/10.1186/2...

>Do we really need AI for anything?

The current research is a valid next step in automation which has always shown to have good ROI imo. I dont think it's very "I" but the algorithms have proven very good at outperforming prior heuristics. this enables automation in areas that have unknowns that cannot be explicitly defined as required by prior tooling. the use-cases it settles into are going to be undeniably better for it, but there is a lot of noise right now with people trying to use it for everything

Pumped storage is certainly the easiest thing to deploy right now, anywhere you have hills. There are projects to make pumped hydro work where there are no hills, but there are deep underground cavities, or deep ocean near enough by you can put a tank under. In those, you pump water up from a space or tank below, and let water back into it to discharge. For the latter, you don't even need a long pipe, just a pump/turbine at the bottom of the tank, so it doesn't even need to be very nearby, just near enough to run a transmission line to. (That applies to any pumped hydro, BTW.)

Ammonia means, simply, synthesizing anhydrous ammonia from environmental nitrogen and hydrogen (that electrolysed from water). To get the power back out, you burn it in your gas turbine, or (in future) use fuel cells. The attractions are that excess ammonia may be sold on the open market, so your synthesis equipment sits idle only when you are actually drawing down on your tank; and it doesn't cost anything more than the tankage to bank it indefinitely. If you need more ammonia than you have banked, you just buy it, maybe shipped in from a solar farm in the tropics.

Remarkably, if you have just enough of some more round-trip efficient storage to get you through the night, the round-trip efficiency of the ammonia, hydrogen, liquified nitrogen, or what-have-you doesn't matter very much. Solar panels are so cheap, you can afford to be wasteful.

Carbon is the 4th most abundant substance in the universe. What is wild is that life is good enough at stealing it away from oxygen to have almost scrubbed it out of our atmosphere. (Compare to Venus.)

It will be important to get the carbon back out of the air, and then not turn around and, as everybody seems so eager to do, burn and exhaust it again. A business model where we end up with the carbon sequestered has been hard to come up with.

There was a suggestion to make it into fiber and put that in cement, thereby strengthening concrete so you don't need as much; but the best source for that carbon is cement manufacture itself, which exhausts CO2 at much higher concentration than in the atmosphere. So, it only helps much if you need a lot more carbon for that than you would have exhausted.

>Remarkably, if you have just enough of some more round-trip efficient storage to get you through the night, the round-trip efficiency of the ammonia, hydrogen, liquified nitrogen, or what-have-you doesn't matter very much. Solar panels are so cheap, you can afford to be wasteful.

i wonder if round-trip efficiency really is all that negligible when solar panel development and shipping is no longer subsidized by fossil fuels and must also use the energy-storage tech of choice as the predominant source of energy for anything not directly hooked up to the grid. a 20% RTE efficiency has a 5x impact on required energy to do anything. Are there enough good locations for solar to cover the needs of all long-term storage needs 5x over? Do we have enough raw materials for that many panels? Seems plausible tbh but idk

In any case, isn't ammonia incredibly toxic? whats the work-around there? we assume perfect handling of it by all parties at every level of a globally large scale operation? A gasoline leak is bad enough, but an ammonia pipeline bursting? that seems like a hard sell

Cost for solar panels is still in free fall. Nothing is being subsidized at anything like the amount we pay for oil extraction: every last barrel pumped is paid out for, cash on the barrelhead, hundreds of $billions every year. You pay for a panel exactly one time, and then it delivers for years.

The point about round-trip efficiency being negligible is that you only rarely need to draw down on those easy-to-ship, easy-to-store media. All you need is, say, 2x more panels to use to top them up again. Most of the time, those extra panels are driving synthesis to sell, generating revenue.

Shipping is always cheap. In the future, we can expect the ships to burn ammonia synthesized in the tropics.

Storage costs are falling even faster than for solar panels.

Solar can be co-sited with roofs, where it extends the life of the roof; with parking lots, where it extends the life of the cars; with reservoirs and canals, where it cuts evaporation and biofouling, and also operates cooler, thus 3% more efficiently; and with pasture and farmland, where it increases yield by reducing heat stress, and cuts water demand. So, yes, there is way more room for solar panels than you will ever want.

Solar panels are usually made of silicon, which makes up 28% of the Earth's crust. The mass of solar panels needed per kW is falling even faster than their cost.

Ammonia is toxic. But we already handle literally millions of tons of it every year without incident. It is lighter than air, so if it leaks, it can't blanket a nearby town like, say (just picking at random) methyl isocyanate.

> There is a lot of value to be earned by closing the gap between cutting edge, but proven, science and easily understood training documents for contract workers.

Oh absolutely! But those are two completely different things.

> I would say most people would not care if we stopped using tax dollars to fund string theory research or finding 'habitable' planets, and a host of other luxury topics.

I agree. And then all of those people will be shocked to see $SOMECOUNTRY do crazy stuff in a decade when the US will be reduced to rubble or decline to irrelevance. Something like that almost happened with Goddard vs von Braun during and prior to WWII. Then, the US suddenly faced rocket interceptors, ballistic missiles, and jet fighters. The whole of US research, and in particular the Air Force, was reorganized after that experience.

> This standpoint is very specific to the world we live in today.

No, it's just very specific to your personal opinion. I take it you wouldn't have funded a patent clerk in Switzerland in 190x when he studied what happens when you sit in a train going close to the speed of light. I'm glad someone in the US listened to him, because otherwise it would have been Nazi Germany to build the atomic bomb that came from his insight.

The whole crux of the matter is that we don't know in advance what pays off.

>The whole crux of the matter is that we don't know in advance what pays off.

Yeah we all know that. My post was sensitive to that. You're not wrong, but there is a cost to that.

>I take it you wouldn't have funded a patent clerk in Switzerland in 190x

Please cite sources on Einstein getting funding for his research from the government. I do not believe that happened, at least not while anything was theoretical. His PhD advisor worked for free, and Einstein published multiple significant papers on his own accord.

He was later paid by the government to take his theory and make the most out of it in reality, which is what resulted in the development of the weapons that won the world war.

So before you dismiss what I am saying for being a bit counter to the typical narrative about science funding - please be aware that what you want is what Germany did in that situation. Which was continue to develop theories of new fringe ideas instead of taking the new golden goose egg that was discovered and making something useful out of it right away

"The simple answer is: funding. There is no (public) money to revisit "known" ideas and results to help clarify them, apply them, or place them in better context. And I would not be the one to explain to the public to spend their money on known ideas rather than discover something new and shiny."

Which is a pity. There's a prominent member of my field who does this for causal inference work, and it has been invaluable.

The problem at heart is research pace. Slowing the pace should help. i.e., it should be ok for faculty to produce half the papers they now produce and still retain their hopes on tenure, or whatever grants and glory.

I cannot get creative under pressure. I don't know how anyone can.

There is insurmountable amount of papers produced every year in AI, and everyone's in a rat race for some reason I don't understand. There is no way someone can exist in a department for more than a year without pushing a paper.

At least in AI, the courses students are taught, professors have an understanding are out of touch with the kind of mathematical rigor needed.

I would guess that the rat race in that case is probably driven by the ability of the machine learning industry to pull people from academia. So the rat race is a literal one, with people trying to get hired by Google, Meta, and the like.

In general though, I am with you. I don’t understand why people are in such a hurry all the time these days. I have a personal principle that I move slow to move fast. Moving slow and methodically is almost always faster in the end.

Reply on news.ycombinator.com