Making Lemonade

I was pretty dour last week.  I am determined to “take that frown and turn it upside down” this week!  Rather than just criticizing DOGE, I have a modest suggestion about how to use that DOGE energy.  And a few short mentions of optimistic progress.

An Alternative Proposal for DOGE

I do not think Elon is dumb — he is very smart.  I may not want to have a beer with him, but there is no denying his effectiveness as an industrialist.  He may not have been the founder or key technical contributor to SpaceX or Tesla, but he has driven those organizations to success.

So, while I am critical of DOGE, it is not because I think Elon is dumb.  I just want to use his considerable talents in a leveraged way on a tractable problem.  And the cost-cutting emphasis of DOGE is worthy but cannot meet the goals it has set for itself.   Paul Krugman describes the US government as “an insurance company with an army,” and if you are politically unable to cut insurance payouts or army expenses, well then, cost-cutting will not address deficits in a meaningful way. 

Accelerated growth is the only politically feasible way to solve our deficit woes.  Imagine if we could reach Satya’s dream of 10% sustained growth — our deficit problems would disappear.  This CBO sensitivity analysis suggests that even a half-point increase in total factor productivity would dramatically affect our debt levels. Whacking headcount is easy — and has little benefit.  Moving the growth needle is hard — but has dramatic benefits, and we should use our best brains to try to move the growth needle.

So how can we move the growth needle?   One of the most commonly identified issues is permitting.  We should massively simplify the permitting process for infrastructure, energy, and other large projects.  Many have written about the complexity of permitting, about NEPA, about the ability of lawsuits to slow down any project.  I used DeepResearch to pull together a lot of this info to review.  The US system is way more burdensome than every other leading economy.  We don’t want to be as authoritarian as China, but we should try to outperform EU nations.

One useful DOGE idea is a publicly available report card — every part of the government should try to have an intelligible online report card.  We should create a permitting grade card — rather than the current DOGE focus of “Total heads” and “Budget dollars saved”, start measuring “days to permit approval” or “days to project completion” for different project types — a large-scale transit project, a significant energy project, an urban redevelopment project, for instance.   Once you start creating and publishing a metric, there is always a powerful drive to improve against it.  Vivek Ramaswamy was purportedly pushed out of DOGE because he wanted to focus on regulation simplification rather than cost-cutting — but he was right; we need to reinject some of this thinking.  

We should also pour any DOGE savings into federal R&D — DARPA, NIH, NSF, and other agency research programs.  We should particularly ensure that leading growth industries are well supported — quantum, space, AI, robotics, electrification, gene editing applications, etc.  Federal R&D spending has been hugely positive for our economy — here is another DeepResearch paper on that, a little repetitive, but supports the core message that federal R&D spending has been hugely effective. 

And let’s have a report card on our R&D spending and our competitiveness in the key industries of the future.  Start measuring and tracking our competitiveness metrics in quantum, AI, space, robotics, life sciences, electrification, materials, etc. How many engineers and scientists hired, how does that compare with efforts in other leading nations, how much GDP have we created, etc.  

We need energetic, enthusiastic outsiders pushing on the government — bringing in Elon and other outsiders to stimulate progress is not a terrible idea.  I want to say yes to that, but I want the energy directed in ways that will have a material effect on our growth rate and, thus, on our deficit, our standard of living, and our world competitiveness.  And we can do it without the haphazardness and casual cruelty of the current DOGE effort.

Short Notes

Construction Physics has a great cheatsheet on energy use and consumption. If you want to think about energy (and tech, as it is an increasing driver of energy usage), this is a great article to help you build your intuition about energy.

Want to build a calculator app?   It sounds so simple, you could do it in a weekend! Or maybe not as this writeup explains — getting math right is surprisingly complex.

It is getting so easy to jam an LLM into any coding project.   It seems like “import llm” will be the first line in every coding project in the future.

This sensor can take any gas and tell you what is in it — it’s like a supersensitive digital nose. Moore’s Law comes for everything.

Microsoft announces a significant breakthrough in quantum computing and maybe a new state of matter.  I started out as an electrical engineer and learned all about circuits and transistors, down to the electron level, and had just enough physics to be dangerous.  It always seemed like moving large collections of slow, heavy electrons around was an incredibly inefficient way to perform computation.   If we can master this technology, it is going to create an unimaginable explosion in computing power.

“AI-designed chips are so weird that 'humans cannot really understand them' — but they perform better than anything we've created” — we on the cusp of an AI-fueled explosion in low-level materials and device design.

Vaccine for pancreatic cancer shows promise — wow, this would be fantastic.