AI, High Explosives, and NIF’s 10th Anniversary

Hello World!


Last Tuesday was the tenth anniversary of NIF! When the laser was built, the plan was to have fusion within five years, and that most certainly did not happen. But according to Einstein, if we knew what we were doing, it wouldn’t be called ‘research’. Happy anniversary NIF!

There were many celebrations to commemorate the 10th anniversary. The first one was held on Monday and it was basically talks about what NIF has done so far. I wanted to go, but I had just been given my new automation project and had to get to work on that. Also, whenever they have big talks about NIF they always over-emphasize the military aspects because sometimes there are high-level government folk in the audience or watching the recording who they need funding from. The federal government is more likely to fund military research than experiments to study how stars work, so they mostly discuss the stockpile stewardship project. I’m much less interested in the military applications so I decided not to attend the talks.

Hmph. What type of congressman doesn’t stop their conversation with the head of NIF to talk to some random intern kid?

On Tuesday however, there was an event with food! Any place that has free food is a place I will go, and also the amazing Mikhail who runs my division invited me. Several congresspeople, Bill Goldstein who runs the lab, Mark Herrmann the head of NIF, John Marchand the mayor of Livermore, and many others were there. I said hello to Bill Foster who’s the representative of the 11th district of Illinois (democrat) but he didn’t respond. It was pretty understandable because he was talking to the head of NIF, and it was quite crowded and noisy. The congresspeople who spoke referred to themselves as the ‘geek caucus’ and I can’t tell how much of that was real and how much of that was pandering to the highly scientific and geeky audience. Interestingly, they seemed very concerned about recruitment. Goldstein, Herrmann, and one of the congresspeople all mentioned that NIF needed new, younger workers, and many of those people are going to Silicon Valley. This explained why this whole summer they’ve been talking to us like we’re going to work here in the future! They’re really trying to get younger people in the lab. They’re doing a pretty good job, but they could be doing better. With scientists, this is one of the best places to work because of the wonderful labs. Programmers on the other hand could be working in many other places and can afford to be extremely picky. I had a buddy who was graduating, and he had job offers from five different companies. He and his friend were discussing which one was most likely to become a unicorn, and at one point they were comparing the nap rooms. Here’s what I’d say about their recruitment in regards to programmers.

What is working:

  • Talking nonstop about how AWESOME the laser is. Most people become engineers because they want to do something meaningful.
  • Holding talks for us about science and engineering with leaders in the field.
  • Giving us lots of free food and swag.
  • Paying us fairly well for young interns.
  • Having projects that look good on your resume.
  • Ping-pong and popcorn every Friday.
  • Hosting a hackathon and giving us equipment for the hackathon.
  • Having Star Trek: Into Darkness filmed at the lab. One of the speakers actually mentioned that this helps them recruit.

What they should do differently

  • The bureaucracy is so slow! We had an entire day of extremely boring training and it took two days to get me admin access. For some people it took longer.
  • Make the programming areas of the lab look more like tech offices and less like something from Hidden Figures. Maybe bring in some gaming consoles, free coffee and snacks? Perhaps the ubiquitous beer fridge? The last one might be difficult because some interns (myself included) are too young to drink, but nearly all startups have those things.
  • Have a better codebase, and use modern programming practices. According to the head of NIF over one third of the US’s nukes date back to the Manhattan Project, but that doesn’t mean our coding style has to. Some of this code was copyrighted before I was born. Most of the difficulty I’ve been having with the code is because of the terrible legacy code I’m dealing with, rather than real technical challenges.
  • Host a table at MLH hackathons! Let the coders know the lab is looking for them, maybe give prizes to people who create NIF related hacks.
  • Emphasize the fusion and not the weapons. Many Silicon Valley folks can be a bit conservative economically speaking, but nearly none are conservative socially. Most of the programming interns I’ve met are a bit at odds with the highly military culture of the lab.
  • Okay this last one is a bit crazy, but with a DoD and DoE budget I think it’s possible since we have access to the funds that flew Katy Perry to Iran to sing for the troops. If you want to get skilled programmers to come to the lab, you need to make Livermore more interesting. I personally don’t care much because I spend most of my time at home programming, but many other people are going a bit crazy because we’re in the middle of nowhere. Palo Alto was boring until the tech boom, and now it’s amazing. Could they do the same thing with Livermore? Just pump some federal money into the town to make it awesome?

Anyways, just some ideas. We also got a nice lunch and these cool rectangles of glass that’s the same glass used for the laser! I got to meet with all the people who worked at NIF, which was nice.

On Wednesday, there was a talk by the head of the National Nuclear Security Administration, Lisa E. Gordon-Hagerty! It was invitation only and I’m an intern, so I wasn’t given a ticket. However, the head of each division got one extra ticket to give out to an intern, and for whatever reason Mikhail decided to give the ticket to me! He said it was important because I was still ‘starting my career’ and it was important that I attend. He also mentioned that the NNSA Secretary was a woman and it would be neat if I could hear her story. Like I said, Mikhail is an awesome guy. Gordon-Hagerty talked about the labs, how they were doing, and what they would be doing in the future. The funding is pretty good because it’s a military facility, but we we’re still behind. Over 50% of the US’s nukes are over 40 years old and over a third are from the Manhattan project, and at the moment we’re not building any big ones. Instead, we’re ensuring that old nukes don’t degrade. She also spoke a bit about how the money was being allocated in the DoE. Unsurprisingly, almost nothing is being allocated towards climate change issues. Grrr……

I asked about how the presidential administration was effecting the lab, and she responded that ideally, the lab would be mostly unaffected by the presidential administration. In reality however, there is definitely some differences between presidents! She never mentioned the president by name, simply calling him “the president.” She also mentioned that he “knows who we are” which I think is a fairly low bar. I’m a bit confused as to how this administration is affecting the lab. She said that the current administration is good for the lab, because they were getting more funding. However, an earlier presenter from the Pentagon said that despite additional funding, many of the projects that really needed to happen weren’t being approved. It was certainly interesting to listen to people who were happy with how the national budget has been split up!

Then I got to tour the High Explosives Applications Facility! It was pretty cool. Most explosives facilities, like Site 300, have the production and testing in different locations. That way if there’s an accident during testing, the people working on production aren’t harmed, and no chain reactions are set off. At HEAF however, they do all in one building! It’s more efficient that way, and they use different methods to keep things safe. Sometimes there’s something that is do dangerous to be blown up in HEAF , and that’s when they send the explosives to Site 300 for testing.


The explosives that they design are made to be very precise. Unlike most explosives, they won’t go off in a fire or a car crash. They are set off only under extreme heat or shock conditions, which are not generally reached unintentionally. By controlling the heat and the shock, you can control the explosion to an extreme degree of precision. In addition, they test homemade explosives to see if they are real threats. If they are, then they program the airport scanners to detect them. They also 3D print little plastic dog bones filled with the explosive to give to bomb sniffing dogs! There’s not enough explosive to actually combust, so the dog is safe.

On Thursday I went to a talk given by Bob Butler who used to work at the pentagon and is now working in a bunch of government security positions. He talked about how the US needs to improve its cyber security, and what other countries were doing. What’s struck me the most from all the military talks is how scared people seem to be. Regardless of whether or not it’s justified, it’s interesting to see how afraid the US military leaders seem to be at the moment.

After work, I’ve been designing an AI that plays tic-tac-toe! It’s a simple minimax algorithm. Basically, after every move, it analyzes all the possible moves by both the computer and the player. In a situation where it wins, it gains points, and in a situation where it looses, it looses points. It then attempts to maximize the the amount of points. At the moment it plays tic-tac-toe, but doesn’t play all that well. I think it has something to do with the way I am storing the points. When I figure out how to make it work (hopefully soon) I’ll talk more about the code.

May you be ever victorious in your future endeavors!