Comment by colincooke | Hacker News Reader

Comment by colincooke | Hacker News Reader

This bill has a ton of what looks like awesome elements inside it, I encourage folks to peruse the summary to explore a little bit. One part that people here may find exciting is the fusion research funding:

> Sec. 10105. Fusion Energy Research. Subsection (a) amends section 307 of the Department of Energy Research and Innovation Act (42 U.S.C. 18645) by authorizing $50,000,000 per year for FY 2023 through FY 2027 for research and development of fusion materials. It extends the authorization for inertial fusion research and development, alternative and enabling concepts, and the milestone-based development program through FY 2027. It authorizes the establishment of at least two national teams to develop conceptual designs and technology roadmaps for a pilot fusion plant, and authorizes $35,000,000 for FY 2023; $50,000,000 for FY 2024; $65,000,000 for FY 2025; $80,000,000 for FY 2026; and $80,000,000 for FY 2027 for these activities. It directs the Secretary to establish a high-performance computation collaborative research program and an associated innovation center in high-performance computing for fusion. It directs the construction of the Material Plasma Exposure Experiment including $21,895,000 for FY 2023 and $3,800,000 for FY 2024 to carry out the project. The subsection also authorizes an upgrade to the Matter in Extreme Conditions endstation at the Linac Coherent Light Source.

What I’d like is that the teams be rewarded with significant monetary renumeration if they achieve a milestone and curtail the subcontracting disease that permeates government purchasing.

> teams be rewarded with significant monetary renumeration if they achieve a milestone and curtail the subcontracting disease that permeates government purchasing

This has a poor track record for basic research and complex technologies. The Ansari XPRIZE recipient’s SpaceShipOne is a prime example.

DARPA Grand Challenges worked well. Created Velodyne Lidar and the team that became Waymo among others.

> DARPA Grand Challenges worked well

Agreed. But the dollar amounts were small. And in many cases, every qualifying team received a grant.

I believe most of the grant money was conditional on meeting significant intermediate milestones.

These sorts of challenges work great for engineering problems where a little bit of money and a lot of sweat is all that is required to figure something out.

Fusion research is not that- it needs very large equipment investments for decades for even the most trivial experiments.

If we had tried to get to the moon with the same model we'd still be waiting!

> engineering problems

This is what is holding fusion from commercial viability.

Give me a modern tokamak with 75 years of known dynamics, increase the size and magnetic field (i.e. bring high temp superconducting wires to market).

At the very least you get something as a consolation prize that has real uses vs. decades of wacky research ideas that more than likely will go nowhere.

IMO thinking of fusion as needing a "silver bullet" to viability is one of the major issues. Everyone wants the sexy new reactor concept, but nobody wants to be improving the boring donut design that's on the edge of practicality.

https://en.m.wikipedia.org/wiki/ITER

“It will be the world's largest magnetic confinement plasma physics experiment and the largest experimental tokamak nuclear fusion reactor.”

Does the $3.9B spent on ITER not count?

https://www.science.org/content/article/cost-skyrockets-unit...

$80 Billion was spent between 2014 and 2017 on R&D of autonomous vehicles [1]. That's 12 Billion / year (& numbers from a while back - I'm sure investment has kept going). The US government spends ~18 Billion on the entire DOE, the majority of which appears to be keeping facilities running / maintenance (even though it classifies all of that as R&D) [2]. Fusion saw about ~4.7 Billion in funding in 2022 [3].

Orders of magnitude matter and fission / fusion R&D has been criminally underfunded for decades.

[1] https://www.brookings.edu/research/gauging-investment-in-sel...

[2] https://www.energy.gov/ne/our-budget

[3] https://www.newsweek.com/nuclear-fusion-report-private-inves...

To be fair, a decent part of the current model for getting to the Moon does involve milestone based payments, with the general idea being that it's cheaper, more efficient and has fewer bad incentives than just endlessly funding until the desired outcome is achieved. IIRC pretty much everything in Artemis besides SLS+Orion is using a milestone based funding model (the lander, the space station, the various research orbiters and landers that started launching this year).

Self driving is a little bit of money and a lot of sweat? Umm... I don't think I've ever seen such a major understatement.

The DARPA challenge kicked off a many many billions of dollars arms race of private capital and we're only just starting to see a trickle of self driving cars on the road some 18 years later. Not to mention meaningful R&D advancements in AI, sensing software algorithms, HW sensing capabilities, etc that have seen applications outside of self driving.

I think NASA's commercial cargo and crew programs were another good example, leading to the SpaceX Dragon and Orbital Cygnus. Of course, the Boeing Starliner's been having issues, but that's why competition is important.

What was the issue with the Ansari XPRIZE? I’m not aware of any controversy about it or the winner.

I’m super happy to see these things, and some money is better than no money, but judging by how much my lab spends per year, $50M/yr isn’t going to get them very far. And we do consumer electronics.

Reply on news.ycombinator.com