The Endless Frontier Act | Hacker News Reader

The Endless Frontier Act

benjaminreinhardt.com

39 pointsrfreytag14 comments

It seems to me as if all the economic criticisms of grant writing and funding committees (it's rent seeking that can destroy value, it distracts people who actually know science, it makes short term bets, it encourages fashions) could just as easily apply to venture capital. What's really different?

Venture capital tends to make members of the elite even richer. Imagine an angel investor putting in some seed money for 5% of the company, then a massive series A from government VC. The angel investors are very happy with this model.

At least grant writing and funding committees are usually at least partly democratic.

Large untouchable and unaccountable pools of capital have the potential to re-create some of the worst excesses of central planning.

I'm not sure exactly what the right analogy is, but although VC is apt in some ways, it's not really appropriate in many other ways.

With research the goal is accomplishing some project. That is the goal. It's not to become a self-sustaining business because that's not the goal, and there's no way to do that anyway. You're not asking to jumpstart something toward independence, it's more akin to VC where you have to return over and over to the same VCs forever. Moreover, there's pressure to ask those VCs for funds even if your project doesn't need their money, or to ask them for more money than you need, because there's some third party pointing a weapon at you so they can skim off funds. In fact, because independence isn't even possible or relevant, things get weird in terms of whether or not you get actual revenue from customers (ala citations, impact, or whatever).

So yeah, it's like VC, but where the VCs don't pay attention to your actual profitability, and instead fund whatever they happen to be interested in, and are really bad at financial management, and you're required to go to them anyway for funds because someone is profiting off of it.

A better analogy in my mind is if everyone hired plumbers not to fix their plumbing, but to apply for government money for plumbing projects because those plumbing grants paid handily to the homeowner. Everyone ends up hiring good grant writers and plumbing in the country falls apart.

Why this focus on funding academia/universities exclusively? All research grants I stumble upon have the requirements of being an official university "researcher".

Are talented enginners not capable in advancing science without a university affiliation?

University research is cheaper than industry because most of the actual work is done by underpaid grad students and postdocs.

I really like OP's idea of giving lifetime annual grants to a lot of promising individuals to use it towards whatever they want. Since they are federally funded, all their work should be eventually open sourced and patents expiring after X years. Universities and private companies can then compete to provide grad students/employees to these individuals in exchange for a fee or other revenue sharing arrangements.

Not only can you save all the time writing grants, it can lead to interesting outcomes like individuals with grants banding together to collaborate on large projects.

I don't know how to setup a repeatable process to identify such individuals though.

IMO, the entire $100B should be allocated to fusion. And then add as much money as necessary until it works. Put together a Manhattan project or something. It does seem like we aren't that far off this time.

Most other things on that list will happen just as quickly, if not more so, without any government help. Fusion is not one of such things. It's also by far the most "endless frontier" thing on that list.

Frankly I in no way understand the utility of fusion. Almost every credible fuel source is difficult to obtain, and for what? The biggest fusion reactor in lightyears, already fueled up for the next 15 billion years, is currently dumping energy onto the planet whether we want it or not. Solar just makes more sense than anything else.

That's the kind of thinking that leads to rolling blackouts and landfills full of solar batteries.

Then instead of 100B on maybe getting to usable fusion, let's spend $XX billion on solving those problems instead.

What would $100b specifically directed into panel conversion efficiency, grid improvements, and battery recycling look like vs the equivalent amount on fusion? I think we'd get a lot farther toward total clean energy conversion.

That said, we absolutely should keep funding fusion research at some level - assuming we can get all the pieces working, it would be a great boon, especially for off-planet power somewhere down the road. However, the next 20 years are probably THE critical moment in stopping the worst effects of climate change, and I think betting on solar+battery tech is more likely to pan out in that timeframe than fusion, especially since we already have the tech to roll out a huge amount of solar panels RIGHT NOW if we wanted to spend the money.

We're past the point of optimal solutions. We're running out of time, and compromises are necessary. Trading catastrophic sea level change in 50 years for panel+battery recycling problems today seems like a fair enough deal to me.

Deuterium-tritium fusion produces neutrons, which can be used to turn lithium into tritium. So in some sense, the required fuels are deuterium and lithium, both of which are abundant enough to be used for fusion.

The sun is a great fusion reactor, but it is also kind of far away. By the time its energy gets to Earth, it's spread out to about 1.4 kW/m^2. The actual power produced per m^2 of solar panels is about 20 times smaller, since the Earth's surface area is 4 times its cross-section, and solar panels are around 20% efficient.

The United States consumes about 3,911,000,000,000 kW h / yr for an average power consumption of 446 GW. By my calculations, an area of 6.37 billion square meters would be required to supply this through solar power. According to Wolfram Alpha, this is about twice the area of Rhode Island. As well as taking up a large amount of land, that is going to require an enormously large amount of materials to make those solar panels, including some with polluting production processes. This is without worrying about storage, or any other details like that.

I like solar power, and I think it certainly has its place, but I hope you see why people also want to make artificial fusion work.

Also, there's a ton of modern approaches to fission that just work and can supply our energy needs for at least another 1000 years with arguably less environmental impact than solar panels.

Agreed, though the question of how much funding do we give to each of the current promising-seeming approaches is a difficult one. Going all-in on a single approach would be very risky, but spreading the funding so thinly that no approach succeeds would also be a mistake.

My personal choice would be to give SPARC (a tokamak design, somewhat smaller than ITER since it would use better superconductors to produce the magnetic field) and magnetized target fusion (see General Fusion for a company that is doing this) enough funding that they can produce working reactor prototypes as fast as possible. Spread the remaining money across a few other promising approaches. I'd even give some small amount to super-risky "tabletop fusion" experiments, just because it's fairly low cost.

Reply on news.ycombinator.com