Dec 7, 2018
My Guest this week is Malcolm Handley, General Partner and Founder of Strong Atomics.
The topic of this conversation is Fusion power - how it’s funded now, why we don’t have it yet, and how he’s working on making it a reality. We touch on funding long-term bets in general, incentives inside of venture capital, and more.
[00:00:00] This podcast I talk to Malcolm Hanley about Fusion funding long-term bets incentives inside of venture capital and more Malcolm is the managing partner of strong atomics. Strong atomics is a venture capital firm that exists solely in a portfolio of fusion projects that have been selected based on their potential to create net positive energy and lead to plausible reactors before starting strong atomics.
Malcolm was the first employee at the software company aside. I love talking to Malcolm because he's somewhat of a fanatic about making Fusion Energy reality. But at the same time he remains an intense pragmatist in some ways. He's even more pragmatic than I am. So here in the podcast. He thinks deeply about everything he does.
So we go very deep on some topics. I hope you enjoy the conversation as much as I did.
Ben: Malcolm would you would you introduce yourself?
Malcolm: Sure. So I'm Malcolm heavily. I found in strong [00:01:00] atomics after 17 years is software engineer because I. I was looking for the most important thing that I could work on and concluded that that was kind of change that was before democracy fell off the rails.
And so it was the obvious most important thing. So
my thesis is that climate change is a real problem and the. Typical ways that we are addressing it or insufficient, for example, even if you ignore the climate deniers most people seem to be of the opinion that we're on track that Renewables and storage for renewable energy are going to save the day and my fear as I looked into this more deeply is that this is not sufficient that we are in fact not on track and that we need to be looking at more possible ways of responding to [00:02:00] climate change.
So I found an area nuclear fusion that is that it has the potential to help us solve climate change and that in my opinion is underinvested. So I started strong atomics to invest in those companies and to support them in other ways. And that's what I'm doing these days
What did founding strong atomics entail?
Ben: and he did a little bit more into what founding strong atomics and Tails. You can just snap your fingers and bring it into being
Malcolm: I almost did because it was extremely lucky but in general Silicon Valley has a pretty well worn model for how people start startups and I think even the people getting out of college actually no a surprising amount about how to start a company and when you look at Fusion companies getting started you realize just how much knowledge we take for granted in Silicon Valley.
On the other hand as far as I can tell the way [00:03:00] that every VC fund get started in the way that everyone becomes a VC is unique. It was really one story for how you start a company and there are n stories for how funds get started. So in my case, I wasn't sure that I wanted to start a fund more precisely.
It hadn't even occurred to me that I would start a fund. I was a software engineer and looking for what I could do about climate change. I'm just assuming that I was looking for a technical way to be involved with that. I was worried because my only technical skill is software engineering but I figured hey, but software you can do many things.
There must be a way that a software engineer can help. So I made my way to The arpa-e Summit in DC at the beginning of 2016 and went around and talked to a whole lot of people if they're different boots about what they were doing and. My questions for myself was does what you're doing matter. My question for them was how might a software engineer help [00:04:00] and to a first approximation even at a wonderful conference like the arpa-e summit.
I think you'd have to say mostly these things are not moving the needle mostly in my terminology. They don't matter and it really wasn't clear how a software engineer could help and then because I was curious because I'd read many things about. Companies claiming that they were working on fusion and they were closed and made an effort to hit every Fusion Booth.
I could find and a one of those booths. I said, I'm a software engineer. What can I do and they said well the next time this guy comes to San Francisco, you should organize an audience and he'll give a talk and won't that be fun? So that guy is now one of my science advisors, but that was. The first part of my relationship there.
So he came I organized the talk we had dinner beforehand and is like how close is fusion and he says well, it could be 10 years away, but it's actually [00:05:00] in infinite time away. And the problem is we're not funded. So then you say well how much money do you need and it turns out to be a few million dollars you say that's really really dumb here.
I am in Silicon Valley my. The company I work for is sauna making collaboration software for task management just raised 50 million dollars in here. These people are credibly trying to save the world and they're short two million dollars. Maybe I can find some rich people who can put some money in the answer was yes, I could find a rich person who is willing to put some money in and Rich.
By and large unless they're really excited about the company do not want to put money in directly. They don't want that kind of relationship. So you work through all the mechanics here and you run as you can convince people to put money in but you need to [00:06:00] grease the wheels by making a normal VC structure in this case.
And then before, you know it you wind up as the managing partner of a one-person VC fund but single investor. And then you say well, I've had a surprising amount of impact doing this. What should I do? Do I keep looking for that technical way to be involved and my conclusion was there's really no contest here.
I could go back to my quest of how how is the software engineer? Can I help climate change or look? I've already put four million dollars into Fusion four million dollars of other people's money, but companies have four million dollars that. Born kinda half without me and several of them are doing way better making way more progress than they would have without me.
And now I have all these contacts in the fusion industry. I can build a team of advisers. I'm in all of these internal discussions about [00:07:00] what's coming next in federal funding programs, and I'm invited to conferences and that kind of thing and it was. So obvious that the way to keep making an impact on climate change was to keep doing what I was doing.
So that ends with my now taking the steps towards being what I call a real VC. Someone who goes out and really raises the next Fund in a much more normal way with multiple LPS and a much more significant amount of money.
Ben: Got it
Malcolm: Ray's right now in the baby if you see they VVC or.
Ben: So you invest in babies?
Malcolm: No. No, I'm the baby. That's and Tina that raises a whole bunch of questions.
Why did you structure the venture as a vc firm?
Ben: So one is why did you decide to structure it as a VC fund instead of say a philanthropic organization if you just wanted to redirect money.
Malcolm: The short answer is [00:08:00] because I can get my hands on way more money.
If this is a for-profit Enterprise, so my all P was very generous and trusting and also very open-minded and part of the four million dollars that I mentioned before actually was a donation. It was a gift to the University of Washington to support Fusion research there because. That particular project that we wanted to support was still an academic project for the others.
The companies were our for-profit companies and there's just no good case to say to someone who has money. You should give money to support these for-profit things in a way that gets you know profit if they actually work you can tap a lot more money if you offer people a profit motive. And I think you create a stronger chain of [00:09:00] incentives.
They are encouraged to give more money. I am more encouraged to look after that money. I have a share of the profit with my fund if it ever makes a profit and
and finally you get a more traditional control structure. I don't yet have. At an actual Equity stake in these companies because we did a convertible note or a y combinator safe, but I sit on the board of the companies. They all know that my investment will turn into voting equity in the future and it's just a much cleaner setup.
So I think there were no downsides to doing it this way and a lot of upsets the bigger question, which I. Contemplated the beginning of all of this was even for for-profit money is a fund the right vehicle or other other [00:10:00] options that I should pursue. That's something that I spent a lot of time looking into it after creating the first fund what other options are there, right?
So one approach is you say, well there are four or however many companies here. I like what they're doing, but they're. Really annoyingly small by Annoying. I mean they are inefficient in terms of how they spend their money and their potentially leaving Innovation on the table. So the companies that I've invested in are all about four people maybe six, but that kind of size and they have one or two main science people in each company those.
Interacts with other scientists a few times a year a conferences those scientists at the conference's are of course not completely trust to love each other. They are all competitors working at different companies [00:11:00] each convinced that they're going to crush the other guys and that's the extent of their scientific collaboration unless they have a couple of academics universities that they're close to.
And when I think about my background in software, I never worked in a team that small I had many more people that I could turn to for help whenever it was stuck. So one thing we looked at seriously was starting a company that would raise a bunch of money and buy these four or so companies. We would merge them all into one.
This is called the Roll-Up. And we'd move everyone to one place. They would certainly have a much larger pool of collaborators. They would also have the union of all of their equipment right? So now when someone had a new idea for an approach to Fusion, they wanted to test instead of needing to contemplate leaving [00:12:00] their job starting a new company raising money buying.
Or scrounging a whole lot of equipment and then yours later doing the experiment. They could practically go in on the weekend and do the experiment after validating their ideas with their co-workers. Right? I think there's a lot to recommend this and it was seductive enough that I went a long way down this path in the end of the the complexities killed.
And made it seem like something that wasn't actually a good idea when you netted everything.
Complexities of Roll-Ups
Ben: Can you go into a little more detail about that? Yeah, which complexities and how did you decided it was not a good idea,
Malcolm: right? So it's much harder to raise money for because you're doing something much less traditional as I guess that's not necessarily harder in some ways.
If you come to the market with a radically new idea. You're so novel that you. [00:13:00] Breakthrough everyone's filters and maybe you have an easier time raising money seen it go both witness. Yeah, and my existing investors was not enthused about this. So I would have certainly had to work past some skepticism there on top of that you have to convince all of these companies to sell to you and that looks really hard.
The CEO of one of the companies told me look I'm a lone cowboy. I think he said and made it very clear that he was used to executing independently and didn't want to be part of larger company. Potentially. I could have bought his independence by offering him enough money that he couldn't refuse but that's not really the way you want to build your team.
Other companies were enthusiastic but it would getting [00:14:00] the majority of these benefits would have required people moving. Yeah people and companies and these companies have connections to universities. Of course, the people have families they have whole lives. It wasn't clear that people wanted to move.
It really looked as if everyone was really excited about a roll up that happened where they lived. Yeah on top of that. These people are cordial to each other at conferences. And at least think they wanted to collaborate more but they're also pretty Fierce competitors. So you also had to believe that when these people were all brought into one company.
They would actually collaborate rather than get into status contests and fights and that kind of thing. Not to mention all the more subtle ways in which they might fail to collaborate and it really big wake-up call for me was when the [00:15:00] two technical co-founders one of my companies started fighting these people had known each other for decades.
They were best men at each other's weddings. They had chosen to found the company together. No asshole VC had bought two companies include them together and force them to work together. This was their choice. And it got to the point where still they could not work together. I went down I spent two days at the company watching the team Dynamic interviewing each person at the company one-on-one and made the recommendation that the company fire one of the founders.
So you look at that and then you're like, well these people say they're happy to cooperate with everyone at these other companies to I really believe that so. Huge caution, I think yeah other people cautioned me that the [00:16:00] competitive factors would be reduced. So I had one guy who went through YC not doing Fusion just a regular software startup say look when we were doing way see we were in the same year as Dropbox and it was clear the Dropbox was crash.
And if we had known that actually we were part of some big roll up and we were going to share and dropboxes success. We would not have worked as hard on our little company as we did wanting to match their success. Yeah.
(Holding companies and how they worked)
So eventually I looked at the third model the first model being the VC fund at the second model being the roll up. The third model was a holding company and this is meant to be a middle ground where we would have a company that would invest in the various Fusion companies that we wanted to support.
They would not be combined. I [00:17:00] guess. I'm neglected to mention several of the other advantages that we would have gotten with the holding with the roll up in addition to a unified team of scientists. We would have had the pool of Hardware that I did mention right we would also been able to have other infrastructure teams.
For example, we could have had a software team that worked on modeling or simulation software that all of the different Fusion teams could use so the idea with the holding company was we would still be able to centralize things that made sense centralist right things where you could benefit by sharing.
But we would have these companies remaining as separate companies. They could raise money from other people if they wanted to or we couldn't provide the money when they needed it. They wouldn't have to move they would be independent companies. But the first thing that we would do is say a condition of taking money from us is [00:18:00] you will give all of your experimental data and enough.
Of the conditions of your experiments to us so that we can run our own simulations using our own software right and match them against your experimental results. We would of course encourage them to use our modeling software as well. But that's harder to force. So the idea was software is something that really can be shared right?
We would encourage them to share it and by having access to their. Their detailed data, we would be able to validate what they were doing and being much more informed investors than others so we could make better investment decisions. We could tell who was really succeeding and who might be struggling or failing so we couldn't make better investment decisions than other investors, which would help us.
It would also help the companies [00:19:00] because our decision to investor to continue to invest would be a more credible signal of success or the value creation and they could use that to to shop it around to raise money from other people. So to the benefits there would be still internalizing some of the externalities while keeping people with their independence, but allowing resource sharing and better signal.
For further support raise so much more flexible sharing sharing where it made sense and not where it didn't and then in an optional way later later on we might have said, well, it turns out that the number of our companies need the same physical equipment may be pulsed Power Equipment, which is a large part of the expense for these companies so we could have bought that.
Set it up somewhere and then said you're welcome to come and do experiments on our facility and you could imagine that over time. They would decide that the [00:20:00] facility was valuable enough that someone from the company moved there. And then maybe they do all their new hiring their and the company's gradually co-locate but in a much more gradual much smoother way than.
In the roll-up where we envision seeing a condition of this purchase is you move right having just talked up the the holding company's so much. Obviously I decided I didn't like that either because that's not what I'm doing one of the death blows for the holding company was doing a science review of the four companies that I've invested in so far.
Plus several other approaches by this point I built a team of four science advisors. We put all of these seven or so approaches past the advisors for basic feedback is this thing actually a terrible [00:21:00] idea and we haven't realized yet or what are the challenges or is this an amazing thing that we should be backing and the feedback that we got was that one of them was?
And should definitely be back right for a bunch of them. The feedback was waiting to see another one. Was it an even more precarious position because of execution problems to more that received favorable feedback did not and still do not have companies associated with them but feedback was positive enough that we.
Pay people to work on them inside basically shall companies so that we own the IP if something comes to that but what did we not so sorry just to interrupt right now. They're in universities right now. They're dormant. They're dormant. Okay a common theme in Fusion is someone does some [00:22:00] work gets some promising results and then for one reason or another fails to get.
Funding to continue that it sometimes the story is then the Republicans got into power and cut the funding or they got less funding than they wanted. So they bought worse equipment and they wanted and therefore they weren't able to achieve the conditions that they wanted but they still did the experiment because of the bad conditions that got bad results.
So they definitely didn't get any more money from that a whole host of reasons. The promising work doesn't continue. Yeah, so in both of those cases there are promising results and no one is working on this got it. Yeah another sad Fusion story. So so bunch of things came out of that science review, but what did not come out of it was oh yes here.
We have a pool of for companies that are all [00:23:00] strong and deserve. And have enough overlap that that some sort of sharing model makes sense on top of that. It was becoming clear that even a holding company was sufficiently novel pitch as to make my life even more difficult for fundraising. Yeah, so it just.
Didn't look like something that was worth taking that fundraising hit for given that the benefits for seeming to be more theoretical or in the future than then in the present
Ben: So with a VC fund to my understanding you are sitting on already given capital and your job is. To deploy it I'm going to use air quotes as [00:24:00] quickly as possible within a
certain limit of responsibility.
Would you ever consider something where you do something there there these private Equity firms that will have a thesis and the look for companies that meet these a certain set of conditions and only then. Will they basically exert a call option on promised money and invest that and it seems like that's that's another structure that you could have gone with.
Did you consider anything like that at all?
Malcolm: Right edit your description of a VC fund and yes, we may I please one is you're not sitting on a pool of money that is in your bank account. Some of the money is in your bank account, but there's a distinction between the money that is committed and the money that is raised.
[00:25:00] So you might say I want to have a VC fund that has 40 million dollars over its lifespan if you wait until you have raised. All 40 million then the deals that you'd identified at the beginning that you are using to support the raising of your fund will likely be gone. It can take a long time to raise even a moderately sized fund.
Yeah, unless you're one of those individuals leading very Charmed lives where in weeks they raise their entire fun, but for the rest of us the fundraising process can be 6-12 months that kind of thing so, You have a first close where you've identified or where you have enough and money committed to justify saying this fund is definitely happening.
Right [00:26:00] we're going to do this even then maybe your first clothes is 15 million dollars. You don't need all 15 million dollars to start making your Investments right now. So you have 15 million dollars. Right, but over the life of the fund you do Capital calls when your account is too low to keep doing what you're trying to do.
Right? So the LPS get penalized heavily if they fail to produce the money that they have committed within a certain amount of time after you're calling it got it you could in principle call all the money at the beginning but you damage your friends metrics if you do that. Got it. Funds are. Graded through their internal rate of return and I remember exactly how this is calculated.
(Internal Rate of Return (IRR) )
But part of that is how long you actually have the money. So if you got the money closer to when you're going to spend it or invest it, you look better got it. So that's the first edit. The [00:27:00] second edit is I wouldn't say my job is to deploy the money as quickly as possible. Mmm. My job is to deploy the money for the best results possible.
I measure results in terms of some combination of profit to my investors and impact on the world. Right because they think Fusion is well aligned to do both. I think these prophets are pretty consistent. So I'm not trying to spend my money as quickly as I can. I'm trying to support a large enough portfolio of companies for as long as I can.
Large portfolio of companies because they want to mitigate the risk. I want to include as many companies in the portfolio. So that promising ideas do not go unsupported. That's the impact and also so that the company that succeeds if when ultimately does is in my fund so [00:28:00] that my investors get a return got it and then I want to support them for as long as I can because the longer I'm supporting them.
The larger return my investors get rather than that later value creation accruing to later investors. Got it. Also. longer. I can support them the greater the chance that the company has of surviving for long enough and making enough progress that it can then raise from other investors investors who probably will know less about fusion and be less friendly to Fusion.
Why not start the Bell Labs of Fusion
Ben: okay, there's there's a bunch of bunch of bookmarks. I want to put there the first thing is one more question about possible structures. So a problem that you brought up consistently is the efficiency gains from having people all in the same place all sharing equipment all sharing code all sharing knowledge that.
Does not happen [00:29:00] when you have a bunch of companies, why are you so focused on sort of starting with companies as or groups of people who have already formed companies as as the basic building blocks. So for example, you could imagine a world where you create the Bell Labs a fusion where you literally just start from scratch.
Hire people and put them all in the same place with a bunch of equipment and aren't working together without having to pull people who have already demonstrated their willingness to go out on their own and start companies.
Malcolm: Yeah, great question and the bell Labs diffusion is an analogy of it gets thrown around a fair amount including to describe what I was trying to do.
Although I agree. It's slightly. I think there are two answers to that question. One is [00:30:00] by the point that I was really considering this. I already had invested in for companies. So partly the answer is path dependence got it and partly the answer. Is that by the time I was clearly seeing the problems with the rollout especially but also the holding company it was.
It didn't seem as if just starting a company from scratch was really going to change that some people make the argument that actually the best plasma physicists aren't in companies at the moment. They are in Academia or National Labs because the best ones don't want to risk their reputation and a great job for a two-bit company that's going to have trouble.
And therefore the if you could come along and create a credible [00:31:00] proposition of the legitimate company that will do well fundraising and prove that it will do well at fundraising by endowing it with a lot of money in the beginning you may then hire those people right? I know some people who are convinced that this is possible.
You still have to deal. The asshole complex that is common with infusion. These people have had their entire careers which are long because they are all old or they're in PHD programs basically to become quite sure of the approach that they want to take for Fusion, right? So it was difficult to find a team of four.
Experienced knowledgeable and open-minded advisors for my science board and not all of those people are able to be hired for any price. I think if you want to actually stock a [00:32:00] company with these people you need more people right and they all need to be able to be hired and you still need to convince them to move and you still need to convince them to work on each other's projects.
So it I think it's an interesting idea. I have real concerned about the lack of competition that you would get about all the areas that I just mentioned and on top of that when I looked into the situation around the software sharing and the hardware sharing more closely. I became less convinced that this is actually available.
What's that on the software side many people don't even believe that it's possible in a reasonable time frame to create simulation software that [00:33:00] can sufficiently accurately simulate the conditions used by a whole range of different approaches to Fusion at the moment. We. Many different pieces of software or yeah codes as the physics Community calls them that they're each validated and optimized for different conditions different temperatures different densities different physical geometries of the plasma that kind of thing.
There are some people who believe that we can make software that spans a sufficiently large range of these parameters. As to be useful for a family of fusion approaches. There are even people who claim to be working on them right now. Yeah, and when you dig more deeply you discover, yeah, they're working on them, but they haven't accomplished as much of that unified solution [00:34:00] as they think they have is they say they have so you talk to other people who use these and they're like, yes.
Yes, I think those people really have the. I think they might be the people who can do this. They're not there yet. So the notion of spinning up a team of software engineers and plasma physicists and numerical experts and so forth to try to do that came to seem like a bigger lift with much more dubious payback in the relevant time frame than I had initially thought similarly on the hardware side.
It is really costly in many ways to reconfigure physical equipment for one experiment and then reconfigure it for another experiment is really bad when you have to move things between locations as well or move a team to a site and configure everything there and then do your experiments for a month, but [00:35:00] it's still bad even.
All the people and all the equipment are in when se you get to the most consistent results. If you can leave everything set up and you want to be able to keep going on Saturday or keep going on Monday because you weren't quite done with those through experiments. So to what degree can you really share these results these certain not these results.
She killed to what degree can you really share this equipment? Yeah, definitely to some degree to a large enough to agree to justify. Spinning up a whole company. I'm not convinced got it on top of that. If I were to start a company doing this, I would need to find a CEO build up a whole team that I don't have to build when I'm investing in other companies, right?
Should I be that CEO of many people assumed that I showed her that I wanted to or something like that. I think it's a really hard sell for [00:36:00] investors that I'm the best person to run this company on the other hand. It wasn't actually clear who should do it.
Incentives: How do you measure impact and incentivize yourself?
Ben: Yeah, that makes that makes a lot of sense. I want to [00:37:00] go back to you're talking about incentives previously both that your incentives are to both have impact and make money for your shareholders.
Yeah, I want to ask first. How do you measure impact for yourself in terms of your incentives you. I mentioned something along the lines of company's existing that would not otherwise exist. So like how it's pretty easy to know. Okay. I've like made this much money. It's a little harder to say.
Okay, I've had this much impact. So how do you personally measure that?
Malcolm: Yeah, the clearest example of impact so far is another project called fuse annoyingly. Annoyingly the same name is spelt differently. So this is the fusion z-pinch experiment Fu [00:38:00] Ze at the University of Washington. And it's the group that we donated to
(Fusion Z-Pinch Experiment. https://www.aa.washington.edu/research/ZaP)
it is all four of the companies that have given money to so far are supported by our pennies Alpha program
(ARPA-e Alpha Program:https://arpa-e.energy.gov/?q=arpa-e-programs/alpha )
its Fusion program and all four of them got less money than. Rpe would have liked to have given them. So the time that I became involved with the fuse project they were behind schedule on their rpe milestones and we made them a donation that enabled them to hire an extra two people for the rest of the life of the project that enabled them to catch up with their milestones and become the.
Most successful of the fusion programs that are P of fusion projects that are PE has [00:39:00] when I say most successful what I mean is they are hitting their Milestones they are getting very clean results. So there they have a simulation that says as they put more and more current through their plasma.
They will get higher temperatures and higher densities basically. Better and better Fusion conditions and that at a certain point they will be making as much energy as they're putting in at a point beyond that they will actually be getting what we call reactor relevant game getting a large enough increase in energy through their Fusion that they could run a reactor off that this and the way we plot their progress is.
We look at the increase in currents that they're putting through their pastor and check that they are getting results that match their theoretical results for them. It's especially clean because they have this theoretical concern this theoretical curve [00:40:00] and their experimental results keep falling very close to that curve.
So it's a really nice story because the connection between the money that they got. From strong atomics and the people that they hired and the results that they were able to the progress they were able to make with those additional people and the scientific validity of what they were doing is clear at every step.
Yeah. So so that's one way that I can see the impact of what I'm doing another way. That's more. Is by being involved in the field and trying to make sure that it all makes sense to me. I wind up having insights or coming to understandings the turn out to be helpful to everyone. So I spent a long time [00:41:00] wondering about the economics of fusion.
Companies are understandably mainly focused on getting Fusion to work and they don't spend that much time thinking about the competitive energy Market that they're likely to be selling into 15 or 20 years from now and what that means for their product. I spend time thinking about that because I want to convince myself that the space matters enough to justify my time.
So I went through the stock process. And came to the conclusion that the ways that the companies were calculating their cost of energy were wrong. They were assuming that the reactors would be operating more or less continuously and they would be able to sell all of the electricity that they made whereas the reality is likely to be that for five or hours or so every day.
No one [00:42:00] will buy their electricity because wind and solar producing cheaper electricity. See, right. So the conclusion that I've come to is well so scratch that so the companies often conclude that they need to be demand following they need to make their reactors able to ramp up and down according to what the demand is right that has other problems because the reactors are so expensive to build and so cheap to run.
That ramping your reactor down to follow the demand doesn't actually save you any money. And so it doesn't make the electricity any cheaper. So I worked through all of this and came to the conclusion, which I think most people in the fusion space agree with know that you actually need to have integrated thermal storage.
Your reactor is producing typically hot molten salts. Anyway, right and rather than turning that into [00:43:00] electricity. You should store the bats of hot molten salt and then run the reactor continuously and ramp up and down the turbine that is used to go from hot molten salts to electricity interesting turbines are cheap.
They have low fixed costs. So you can much more affordably ramp them up and down plus if you were going to be demand following you were already going to be ramping your turbine up and. All I'm saying is keep the turbine demand following right make the reactor smaller so that it can run continuously.
Right which is the most efficient way to use a high Capital cost good and then have a buffer of molten salt. So that's the that's a kind of insight that have come to by working through the economics and overall the investment case for Fusion. That I hope will help all the companies not just the ones I'm [00:44:00] investing in.
Incentives for LPs
Ben: so those are those are your incentives is that combination of impact profit and you also have LPS because of the VC fund structure. Where do you see there? What are their incentives in terms of what they want to see out of this? All right out of your firm
Malcolm: my current LP is anonymous and so there's a limit to what I can say about their incentives sure, but they care about climate change.
They basically by into my argument that climate change is real and worth mitigating, right and.
Fusion is a promising and underinvested potential mitigation.
Does the profit motive increase impact?
Ben: And to go a [00:45:00] little farther into that this is just a comment about impact investing as a whole so the question is could they get a better return? By tape putting that money into a this definitely putting you on the spot, but I think you could probably make an argument that they might get a better return just on the money putting the money in some other investment vehicle.
And so they probably want to see that same impact that you want to see and. I guess the the thing that I'm interested in is does having the profit motive actually increased impact and if so how
Malcolm: regarding the potential for profit. When [00:46:00] I started doing this, I thought it was really a charity play. I guess more politely only in Impact play but set up in a for-profit structure so that if it happened to make a profit then the people who had enabled it to happen would be able to share in that profit right as I have.
Looked at the space more closely and refined my argument or arguments in this area. I have come to believe that there's a meaningful potential for profit here. This is all hinges on what you think on the chance that you were saying for Fusion working. It's very clear that if you shouldnt Works in a way that is economically competitive.
The company that gets there will be immensely valuable assuming that [00:47:00] it manages to retain and say p and that kind of thing. So I've taken stabs figuring out the valuation of one of these companies the error bars are huge. So I got numbers around 25 billion for my low-end valuation closer to a trillion for the high-end.
It's really hard to say, but the numbers are big enough on the profit if it works side right that it really boils down to do I think it's going to work that is a hard thing to put numbers on but by investing in a portfolio of them you increase your chances.
Risk, Timescales, and Returns vs. Normal Firms
Ben: something I know from other VC firms is that.
You have to they have to limit their risk. They don't make as risky Investments because of their LPS because they feel they have this financial duty [00:48:00] to return some amount to their LPS in a certain amount of time right? I do you worry about those same pressures. Have you figured out ways around them ways to extend those time scales.
Malcolm: I don't think I'm going to be subject to the same pressures because anyone who gives me money is going to be expecting something very different. Yeah, so instead of being subject to those pressures. I think that the same psychology manifests for me as limiting my pool of investors. So it's a real problem.
It just plays out differently. Got it.
That's it. A few things are different for me because I'm pitching the fund differently a normal fund cannot look at a space and say it's really important that something works in this space, but [00:49:00] it's not clear which company might succeed because there's real science risk, right so normally. Normally the investors to Silicon Valley can decide that this company or these two companies should be the winner and they can all agree that they're going to put all their money in there and they can anoint a winner it will win because it's getting all the money shorter than major Scandal that does not work for in when investing in companies with a heavy science risk.
That's why I think you need to invest in a portfolio of companies and a normal fun has trouble doing that because they are obliged by their investment thesis that their investors have signed off on to spread their money out across different different sectors. Okay. So again, that doesn't make life.
Magically easy [00:50:00] for me. It means I need to find investors who are on board for doing something different specifically investors who are wealthy enough that they are diversifying their Investments by investing in other funds or Vehicles besides mine and are not expecting diversity from me. But having found those investors, I will then be in a much better position because I can concentrate.
In one sector and really solve that or at least strongly supported
Is Money the limiting reagent on fusion?
Ben: got it that makes a lot of sense. I want to shift and talk about Fusion itself. Okay a little bit more. So I'm sure you've seen the the fusion never plot. I'll put a link up in the show notes.
(Fusion Never Plot: http://benjaminreinhardt.com/fusion_never/)
So the question is this plot makes it look like if you pour more money in it will go faster. Do you think that's actually the case or is [00:51:00] there something else limiting the rate at which we achieve Fusion
Malcolm: if you have to leave the existing spending.
Then adding money is a way to make it go faster. But a cheaper alternative is to spend your existing money more wisely the world's Fusion spending and America's Fusion spending to a first approximation all goes in to eat. The international thermonuclear experimental reactor this International collaboration in France.
(ITER - International Thermonuclear Experimental Reactor. https://www.iter.org/proj/inafewlines)
This thing came about because the next step for a fusion experiment in America and Russia was too expensive for either country to pursue independently, even though everyone's first inclination was surely to keep competing. So became a collaborative [00:52:00] Endeavor and. It's now a collaboration between many countries that things expected to suck up 20 billion or more and has a depressing schedule that ends with Fusion Energy on the grid and 2100.
Okay, America puts on the order of a hundred fifty million a year into either directly and say 500 million a year. Into what are called either relevant projects domestic projects where you're trying to learn about something some problem that's relevant either but you're learning in a way that is smaller cheaper better controlled right than a 20 billion dollar massive building where everything is inevitably really complicated the.
Other placed in America spends money is on [00:53:00] Neff the national ignition facility, which is really a weapons research facility that is occasionally disguised as an energy research facility. Another way that it's been described to me is the perfect energy research facility what these cynical people meant was it's too small to.
(NIF - National Ignition Facility: https://lasers.llnl.gov/)
Actually get to ignition or energy Break Even but it's big enough that the people working on it can tell themselves that it might get there. If only they work harder If Only They dedicate the rest of their career to this. So it has large numbers of people who really care about Fusion Energy much more than bombs working on.
Because it's the best way that they can see meaning the best funded way that they can see to get there but they don't actually seem to believe that it's [00:54:00] going to get there. They just don't have any choice. So we spent a lot of money on these two programs and that funding would be more than adequate for forgetting to.
If we spent it on anything more modern it is not controversial to say that these techniques these two facilities are the best Fusion approaches and experimental setups that we could come up with in the mid-80s the mid-90s when they were being designed that's a fact that's when they were being designed.
They've had limited upgrades since but yeah, that's. That's the overall story. What is controversial is weather continuing to support them is the best move. There are people who believe that we need to keep putting money in there [00:55:00] because we're going to learn a lot if we keep doing that science or because if we don't put the money in there, then the money will get pulled and probably stand on bombs or something like that, but it won't come to fusion and so.
Better bad money in a fusion than worse money somewhere else, right. My personal view is that eater is such a ridiculous energy project that it harms the entire Fusion field by forcing people to pay lip service. To the validity of its goals that we would be better off admitting that that thing is a travesty and that there are better ways to do Fusion, even if it meant losing the money now, I'm not certain about that.
But that's the gamble I would take
good news is we probably won't have to take that Gamble and it [00:56:00] looks as if the federal government is becoming much more open to to a yes and approach to funding. The mainstream approaches diffusion if and eater and a variety of projects for alternative approaches and more basic Research into things like tritium handling tritium breeding hardening materials to deal with high-energy neutrons.
Lasting longer in the face of high-energy neutrons that kind of thing. So I think there's real momentum towards building a
in inclusive program that can support everyone and that is of course the best much as I would take the gamble with killing eater and killing them if they do produce real scientific results, and and if we can have all these things that's a wonderful out.
Government Decision Making and Incentives
Ben: on that note who ultimately [00:57:00] is the decision maker behind where government Fusion money is spent and what are their incentives?
Malcolm: This is America. Is there ever one person who's the decision maker about some
Ben: maybe not one person but is it is it Congress is it unelected officials in? Some department. Is it the executive branch? Do you have a sense? Is it some combination of all of them?
Malcolm: The money flows through the department of energy a sub-department of the doughy is rpe, which has its 30 million dollar Fusion program and will hopefully have a new and larger Fusion program in the near future.
There's also the office of Fusion science in an office of Fusion Energy and science ofes that funds a lot [00:58:00] of the mainstream Research into Fusion
( Office of Fusion Energy Science https://science.energy.gov/fes/ )
arpa-e is to my knowledge created by Congress and fairly independent of the doughy, but there's still feuding. I think without describing malice to anyone. It is a great Testament to many people's conviction and political skills that they were able to get America to fund niff and either more or less consistently over decades at a high cost and.
Those people are highly invested in those projects continuing. I don't know whether that's because they genuinely believe that that's the best way to spend the money or fear that the money would disappear from Fusion completely [00:59:00] if it stopped or don't think that the Alternatives actually have any scientific credibility or.
Are so now trapped by the arguments that they've been making strongly and successfully for decades, but for one reason or another or many reasons,
they strongly believe that we need to continue to do these so there is a tension between people who want to fund the Alternatives and the people who want to fund the mainstream fusion
Who are the government decision makers?
Ben: and who are these people do you have any sense of actually who they are? Like I'm not asking you to name names, but like what is their role?
What is their nominal job title?
Malcolm: I think it's a bunch of civil servants within the Dewey Congress has a role like cotton Congress gets to decide how much money to provide and that's often attached to a. About how that money will be spent right there have been [01:00:00] Congressional hearings on Fusion the covered either and whether we should continue to fund either it's a lot of different people.
What are the roles of Academia, Government, Industry, and Philanthropy?
Ben: Okay? Yeah. I'm just really interested in dating you down until I like where we're the incentive structure is set up along those lines in in your mind in sort of an ideal world. what do you see the ideal rolls of the the four Columns of Academia government private investment and philanthropy in making sort of an epic level project like Fusion.
Malcolm: There's a ton of room for government support on this. The federal government has National Labs that have the best computers the best software which is often classified the best testing sites many in many [01:01:00] ways. The only testing site. And lots and lots of experts the one thing that the federal government lacks is a drive to put Fusion Energy on the grid as quickly and as commercially successful as possible.
I don't rule out that the federal government could develop that drive but. It seems like a long shot given that there's a lot of disagreement about climate change and energy policy and that kind of thing. So I think that the ideal would be that the federal government supports Fusion research with all of its resources Financial expertise modeling modeling software.
Modelling hardware and testing facilities in partnership with Private Industry. So [01:02:00] that Private Industry is providing the drive to get things done. So I imagine a lot of research done at the federal government so that if the current crop of companies bottom out if it turns out that their techniques don't work.
We have more Fusion research coming down the pipeline to support a later crop of companies, but we would have companies working closely with the federal government to try to build reactors getting assistance in all those ways from the federal government and providing the drive. The company's would have this.
Call of Fusion Energy on the grid that they would be working towards but they would get to use the federal government's resources for the areas that they're focusing on. There are also the areas that the companies are not focusing on areas that are largely common [01:03:00] to all companies and therefore no company views it as on their critical path to demonstrating reactor relevant gain, For example, for example tritium is toxic to humans and difficult to contain.
It turns out even hydrogen is difficult to contain it leaks through metal surfaces, but we don't talk about this because hydrogen is. Astonishingly boring and section in small quantities. So we don't care that it leaks out of our containers we do care when tritium leaks out of containers because it's heavily regulated and toxic right?
So any Fusion company that's handling Trillium is going to need a way to contain tritium with very low leak rates. Also, the world does not have very much true. You can't actually use tritium as a [01:04:00] fuel for Fusion. You have to breed tritium in your reactor from lithium. So the real inputs to the reactor if it is a deuterium tritium reactor will be do tarian and lithium and you'll be breeding tritium from lithium in your reactor.
So we also need to study how we're going to breathe. The tritium right? We're making mathematical calculations about the tritium breeding rate how much tritium we will get out after doing Fusion relative to the amount of tritium. We had before doing fusion and these tritium breeding rates are close to 1 if they're below 1 or really not enough above one.
We're screwed, right? So there's important work for academics. And the federal government to do to better understand trillion breathing rates [01:05:00] and what we can do to increase them and write how to make this work.
Companies aren't incentivized to look at things on critical path
Ben: Right and at the companies are not incentivised to look into that right now because they don't feel like it's on their critical path
Malcolm: investors Pope maybe including myself have made it clear to these companies that.
What they will reward the companies for is progress on the riskiest parts, right? This is valid you want to work on the Unruh Burning Down the biggest risks that you have, right and everyone perceives that the biggest risk is getting through Fusion conditions. Sorry. In many of these companies are already getting to Fusion conditions, but everyone perceives that the biggest risk is getting Fusion to work getting reactor relevant games from Fusion, right?
So compared to that these risks are small and it's [01:06:00] valid to the further, right? If you're a single Company If you're looking from the perspective of a portfolio, which the federal government is best positioned to do which is. Going to be somewhat well position to do then your risks are different you're willing to say I have a portfolio of these companies.
I don't care which one succeeds I'm doing what I'm doing, assuming one of them will succeed. Now. What can I do to D risk my entire portfolio, right? You look at it differently and then these problems start to seem critical. So with my second fund one of the things I want to be able to do. Is support academics or maybe for-profit companies that are working on this but the federal government isn't even better fit for this it is perfectly positioned to do this.
What does the ideal trajectory for Fusion look like?
Ben: I think a good closing question is in your in your Ideal World. would Innovation infusion come into being what would the path look [01:07:00] like? Imagine Malcolm? King of the universe and we started with the world we have today what would happen
Malcolm: I think the federal government would do the heavy lifting but it would rely on private companies to really provide the drive. It would the federal government would also support the longer term. Things that are critical but not the highest risks such as the tritium issues today - and perfect.
Why do so few people invest in fusion
Ben: there anything that I didn't ask that I should have asked about
Malcolm: one question that comes up quite often is why so few people invest in Fusion? Yeah and why it is. That I'm the only one with the poor.
Ben: Yeah, if you could if it's there's a possible payoff of a trillion dollars, right?
And that's even if it takes 20 years the AR are still pretty good.
Malcolm: Yeah, the way I think of it [01:08:00] is there's a funnel where. It's like a company's fun offer acquiring customers. But in this case, it's an industry's funnel for acquiring investors and investors are falling out of this funnel at every stage. The first stage is of course, you have to believe in anthropogenic climate change, but we have lots of investors who believe in.
Why do you need to believe in climate change to fund fusion
Ben: quick question. Why why do you need to believe in that in order to want to fun fusion?
Malcolm: Okay, that's a fair point. You don't have to but it's the easiest route. Okay. If you don't believe in climate change, then you have to believe purely infusions potential to provide energy that will be cheaper than fossil fuels, right.
I believe a zoo that but it is a higher bar [01:09:00] than I believe the climate change is going to encourage people one way or another to put a premium on clean energy. Got up when I do my modeling. I'm not taking into account carbon taxes or renewable portfolio standards, but it's nevertheless easier to convince yourself to care about the whole thing.
If you think that this is an important problem, right? Otherwise you could do it just because you think you can make a whole ton of money, but it is a high-risk way of making money, right? So one way or another let's say you decide you're interested in well. No, I think that's carry on with the.
The funnel for a climate change. Yes. Yes, because there are a few more places that they can fall out and these places might apply to an ordinary profit seeking investors as well, but it's less clear. So you've decided you believe in anthropogenic climate change [01:10:00] and you'd like to see what you can do about it.
Maybe you then narrow to focusing on energy. That's a pretty reasonable bad energy is something like 70% of our emissions when you track everything back to the root. So perfectly reasonable Pace to focus within energy. There are lots of different ways that you might think you can do something about it.
There's geothermal power. There's title power a long range of long tail a large range of long tail. Ways that you can make energy and a lot of people really get trapped in there or they decide they're going to look at the demand side of energy and think about how they can get people to drive less or insulate their buildings better or whatever.
And in my opinion most of these people are basically getting stuck on [01:11:00] things that don't move the. Yeah, the don't add up to a complete solution. They are merely incrementally better. So when you look at things that might move the needle on energy, I think that the supply side is way more promising because the demand side is huge numbers of buildings cars people whose habits need to change on and on if you can change the supply side then.
It doesn't matter as much if we are insulating our houses poorly driving too much that kind of thing. If we have enough energy we can doesn't matter we can make hydrocarbons. From the raw inputs and energy and we can continue to drive our cars and flat our planes and heat our cart heat our houses that have natural gas furnaces and that kind of thing.
Ben: Well, if you have if you have absurd amounts of energy, you can just literally pull carbon out of the air and stick it in concrete right?
Malcolm: That's [01:12:00] the other thing. So, I believe that focusing on energy Supply is the right way to go because. With a distributed Grid or not. We have many fewer places that we make energy than where we consume energy and what you said, we're looking at significantly increased demand in the likely or certain case where we need to do atmospheric carbon capture and sequestration.
It takes the energy it takes energy to suck the carbon out of the atmosphere and it takes probably. More energy to put it into a form where we can store it for a long time. Right? So I think you need to focus on the energy Supply even within that we have lots of ways of making clean energy that aren't scalable and there's a wonderful book called sustainable energy without the hot [01:13:00] air
(Sustainable Energy without the Hot Air https://www.withouthotair.com/ )
that catalogs for the United.
All the ways that they could make energy renewably and all the ways that they use energy and is more or less unable to make the numbers add up without throwing new clear in their got it nuclear or solar in the Sahara and then transfer transmitting the energy to to the UK. So that's it. You can probably make Renewables work.
If you can start but the problem with Renewables is that well Renewables come in two forms that are the predictable forms like Hydra where we control when we release energy and then the variable Renewables like wind and solar where they make energy when Nature cases for those. We need something to match the demand.
That humans have for energy [01:14:00] with the supply that nature is offering and you can try to do that in some combination of three ways. You can over build your Renewables to the point where even on bad days or in Bad seasons, you're making enough energy. You can transmit your energy. You can overbuild you can space shift your energy.
By building long distance transmission lines and you can time shifts your energy through storage. Right and it turns out that all three of these have real costs and challenges storage is an area where people get pretty excited. So this is the next point that people fall off. They say wind and solar are doing great. storage isn't solved yet, but gosh it's on these great cost curves and like I can totally see how storage is going to. A solved problem and then Renewables for work and I think there are two problems with that is that these one is [01:15:00] storage actually needs to get a lot cheaper. If you want to scale it to the point that we can use it for seasonal storage.
Right? And we need a solution to the seasonal problem. There are places like California the. Vastly more Renewable Power sometimes if you're in the other times of year in California, it's by a factor of 10 or 12. Wow. Yeah, so it's not a day by day. It's Yeah by month and and that's critical because if you are cycling your storage daily just to to bridge between when the Sun's shining and when people need their power you get to monetize your story.
Every day you get the entire capacity of your stories roughly right used 365 times a year for the 20 years that your plant is left. Right? If you're cycling it seasonally meaning once a year you get to sell 20 times [01:16:00] the capacity of your storage rather than 20 times 365 times the capacity of your storage.
So your storage needs to be one 365th of the price. To hit that same price Target. Yeah. This is a really high bar for storage is economics. So the first problem with betting on storage is it needs to drop in price a lot to really solve the problem. The second problem is lots of people are investing in storage.
So if you want your money to make a difference in terms of making you a profit or having an impact on the world. You need to out-compete all those other people working on wind and solar and storage. you're going to play in that space
and that leaves something like nuclear. There are compared to Fusion plenty of people investing in various nuclear fission approaches. So again, your company can make a difference your money can make a difference there. [01:17:00] But and it'll be a bigger difference than in wind and solar and storage but a small difference than Fusion.
So say you get all the way to I think I want to invest in Fusion. what you now encounter is an industry where everyone you talk to will tell you that their approach is definitely going to work unless they're being really nice and that everyone else is approach is definitely not going to work \so.
Pretty understandable at that point to give up on the whole Space. So in this is the step before that you have to to get over the hurdle where Fusion is always 10 years way. It's been 10 years away for as long as most of us have been alive, you have to decide that you even want to look at Fusion
then you hit the problem where everyone says nasty things about everyone else to a first approximation if you spend a long enough time there you might find a company. The convinces you that they are different that all those other people are crazy [01:18:00] and they're worth investing in but I've talked about why investing in one company is a bad idea to invest in multiple companies.
You have to find some advisors or just be Reckless enough that you can so you have to find. Enough advisors who are open-minded or you have to be sufficiently Reckless to just roll a dice on these companies or maybe just sufficiently Rich that you're going to roll the dice on these companies, but it's a real hurdle to find advisors who are experienced and credible and open-minded enough to support investing in a portfolio of these companies.
Yes, even then, If you're a regular VC fund which you probably are if you have enough money to do this, right we have the problems that I mentioned earlier where your Charter is to invest in a diverse [01:19:00] set of companies and you can't put enough money in these companies to infusion to support a portfolio of companies. so that is the funnel did ends with as far as I can tell just strong atomics coming out the end as the only fund supporting a portfolio of fusion companies. That's why in my opinion. There are not more people investing in Fusion in general along the way.
I got a lot out of this conversation here are some of my top takeaways. There are many ways of structuring organization that's trying to enable Innovations each with pros and cons that depend on the domain. You're looking at Malcolm realize that a VC fund is best for Fusion because of the low return from shared resources and the temperaments of people involved just because there's a lot of money going into a domain.
It doesn't mean it's being spent well. I love the way that Malcolm thought very deeply about the incentives of everybody. He's dealing with and how to align them with his [01:20:00] vision of a fusion filled future. I hope you enjoyed that. It's like to reach out. You can find me on Twitter at been underscore Reinhardt.
I deeply appreciate any feedback. Thank you.