Building a defence startup | Hacker News Reader

Building a defence startup

medium.com

53 pointspossiblelion66 comments

This trend to try to make war more like startups is pretty disgusting. You notice this with the a16z "American Dynamism" war making arm, "move fast and break things" now means create weapons to terrorize people all over the planet for cash. There needs to be resistance to this movement from everyone in the industry.

Most of the most important innovations in history (especially in the hardware and electronics space) were motivated by increasing defense capabilities, so although its great to pretend that humans can coexist peacefully, that is not the reality...and this also ignores the fact that many of these technologies can have positive impact on the world outside of the defense space, once they mature.

Examples : ARPANET, semiconductors, radar, GPS, etc, etc, etc

I'm sure mankind and fish can coexist peacefully at least. After all, Bush would never lie.

A very fine sentiment to have while you live in safety under the umbrella of the most powerful military the world has ever known. However, consider the defence startup scene in Ukraine right now, for example. They are truly about moving fast and breaking things in an effort to protect their country from annexation, assimilation and (for those who don't go along) genocide.

So, ironically, only when these weapons are there and have done their job do people appear, safe from birth till adulthood in their peaceful affluent little suburbs, who issue blanket condemnations of the defence industry as evil horrible people who produce weapons for the express purpose of "terrorizing people all over the planet for cash".

"We sleep soundly in our beds because rough men stand ready in the night to visit violence on those who would do us harm." - no one, probably [1], but the sentiment still stands

[1] https://quoteinvestigator.com/2011/11/07/rough-men/

Defense contracting is one of those things where it’s going to get done with or without you. No amount of protesting will stop it bc the govt won’t boycott defense companies.

Would you rather those companies be startups whos employees are potentially more aware of the impact of what they’re building or just some big company that’s been making weapons since the MIC started?

That’s the real decision imo.

“If I dont do it, somebody else will” is a morally bankrupt and demonstrably false philosophy.

There are limited people with any set of skills on the planet, and infinite things to be done. Where we choose to put our effort and attention determines what gets done.

Its possible that if there is enough demand for something, someone else will do it, but no guarantee that it happens at all.

> Would you rather those companies be startups whos employees are potentially more aware of the impact of what they’re building or just some big company that’s been making weapons since the MIC started?

I don't care what the mindset of the people making the weapons is, the end result is the same. If you help build the war machine, you are personally culpable for the damage it does.

There is a limited labor market, our skills are in high demand. Protesting the creation of military tech can most definitely have a very real impact.

That’s totally fair.

Unfortunately I don’t think protesting the military industrial complex does anything at all.

Can you get literally every engineer or would-be engineer to agree not to make weapons given an enormous personal upside?

It’s a prisoner’s dilemma. Someone will always take the paycheck and if _anyone_ does than the entire protest effort would’ve been for not bc the weapons still get made.

That's not even the best reason. Violence is the guarantor of last resort for our choice of way of life. On long enough timeline this has historically always come down to who can do the most violence and there is nothing that implies this has or will change in the future (just a relatively long great power peace that has allowed more naive people to think the above is not true.) As a democracy with a relatively strong personal liberty slant i am pretty keen on that continuing to exist so i am pretty happy when i see companies trying to breakup the old and busted military contracting system.

Surely you understand this perspective to be one of many in the plurality of worldviews.

I understand that there will always be people who want to create war for power and profit. Just like there will always be people who push for peace and not destroying the environment (c.f. depleted uranium). I happen to have skills that warmongers would find highly valuable but will not deploy them for military use. I will also advocate for others in our industry to do the same.

To be fair, it is one of the few domains where "move fast and break people" is a perfectly acceptable strategy. After all, a good number of the people you plan to break are on the other side.

Do unto others before the do unto you.

In the UK a useful route is via initiatives such as the Defence Accelerator [0]. This will help you bring a product into use but isn't a VC equivalent. It will often be necessary to team with a traditional Defence player, e.g. to get access to subject matter experts. Compared with normal defence procurements (whose timescales can be decadal) these research frameworks are positively speedy.

Often the desired output is to raise the perceived Technology Readiness Level of a product or technology, which in turn translates to reduced risk if the product is used in a mainstream project.

[0] https://www.gov.uk/government/organisations/defence-and-secu...

In the U.S., the best way to go to market as a defense startup is the SBIR program (or similar small business R&D set aside). The second best option is to subcontract to friends at larger defense contractors. Both are excellent pathways.

I know a number of folks who get or have gotten SBIRs. The rules of the game requiring knowing them well in advance of showing up because timing matters.

Knowing the people or knowing the rules?

I'd like to know too but it reads as if they meant knowing the rules.

Mostly about knowing the rules, but obviously it doesn't hurt to know the people.

It's a mixed bag. You obviously need to know the rules, but they aren't too complicated. SBIR topics are released several times a year in huge batches. Pretty much every branch of government has topics out for proposal, military, health, education, etc. Topics vary in specificity from "Pitch us anything" to "We want this very complicated component that does XYZ, is this exact size, and meets these exact specifications and performance values". The latter are usually pretty clearly intended for a specific company who either already makes or is already researching that component, but they're using the SBIR fund for money so they can't just pick them specifically. I can't prove it, but those just seem to be someone that knows someone important and managed to get a very specific SBIR written for their thing. Some are also clearly written by people who have no idea what they want. I used to work for a start up that did a lot of SBIRs while trying to get off the ground. They'd write like 10-15 proposals per year and maybe get approved for 1-4.

That is about the right hit rate based on my experience, too. I co-wrote a lot of SBIR proposals, and our success rate was about 10-20%. But it’s worth it! $100k for the Phase 1, and close to $1 million if you are successful and you are invited to submit for the Phase 2.

No mention of the cost of regulation? Getting compliance with federal and DoD standards can take a small team and months of work, even if you hire a consultant to walk you through it. The investment just to get to the table is high, although you can try to walk the dangerous line of paying lip service and hoping nobody calls you on your shit

> Estonia based startup

But also, publicly complaining about your clients is not a good move for a startup.

I think it depends on the stage you're at. I have helped companies get FedRamp, soc 2 type 1 and 2 and into govcloud marketplace in less than 6 months. I would argue the more expensive part is proving compliance rather than becoming compliant, especially if it's an early startup with a small team and without too much written in stone.

Oh sure. But if the startup doesn't know it's gonna take $MONTHS and $DOLLARS and $HEADCOUNT of non-business-value engineering just to get compliance, that's money and time they may not have

One of the cases where fpunder qualification should absolutely include industry experience.

The ability to torpedo your entire business by signing the wrong requirement is unparallelled for this reason. Indeed, it's often used maliciously.

Here. Pour a drink. Let me tell you a story.

UltraWag, a supplier, signs a contract from MegaWing, a big aircraft OEM, for 2000 UltraWag SuperRadio units at 12k per unit. Under the contract deliverables, MegaWing gets all UltraWing tech data in this specific BIN format. UltraWag business development signs up for this 24M contract, thinking, "we can make BIN all day long, no problem".

Yeah, they should have asked someone with a technical degree.

The UltraWag dev group dips its toe into the BIN spec, and sees that it's impenetrable. They get on the horn with the reps at MegaWing.

MegaWing says, oh, yeah, of _course_ you need help with getting MegaBIN spooled up. Let's get you set up with the MegaWing MegaBIN SDK, the MegaWing MegaBIN IDE, and the MegaWing MegaBIN build tool. That's 1.8M worth of software, so discount the SuperRadio unit cost by a thousand bucks or so. UltraWag signs off, and on they go.

Now, this doesn't solve a lot of the problem. The MegaWing SDK needs configuration, the build system only works for OS2/Warp. MegaWing sends some experts down to UltraWag, 5 consultants at 100/hr for two weeks. That's 40k, another hundred bucks off the unit price. This happens again, with an SDK change and an IDE change.

And now, we've got a problem.

The problem here is that the SuperRadio 12k unit cost has a margin of 12%, or $1440. The initial MegaWing software discount already took a huge chunk out of that, and after a very few consultations the SuperRadio is now getting sold at a loss. UltraWag owes MegaWing money for giving them product.

UltraWag might confront MegaWing about this, to which MegaWing might respond[1], "Sure, but now you have 𝐌𝐞𝐠𝐚𝐖𝐢𝐧𝐠'𝐬 𝐈𝐧𝐝𝐮𝐬𝐭𝐫𝐲-𝐋𝐞𝐚𝐝𝐢𝐧𝐠 𝐌𝐞𝐠𝐚𝐁𝐈𝐍 𝐒𝐨𝐟𝐭𝐰𝐚𝐫𝐞 to use as you please!"

Sure, you have these new tools . . with the MegaWing SDK, the MegaWing IDE, the MegaWing build, the MegaWing everything. It can't be used for anything else. In effect, MegaWing just paid for a product with dollars that can only be spent with MegaWing. This is not a transaction. It's colonialism.

[1] Chock full of their own importance to the cosmos.

This was very interesting.

> For example — we at SensusQ are building a piece of software that can ingest a large amount of disparate information pieces, be that drone images, satellite pictures, social media posts or text documents.

Congrats!

Do you have an article about how you got into it?

What are some startups in your area that are software?

What newsletters do you subscribe?

I would love to get more in the loop.

If you’re solely in the defense contracting space for the money, there are easier paths to take.

1. Get job at a defense contractor, get a high level clearance

2. Be invaluable to your government client. Tech skills are good but also having a good relationship with the client and having an eye for business needs is even better. The bar is low here.

3. Once you’re invaluable become an independent sub-contractor. Can flip your job or find another gig and collect most of the bill rate. If you make 150k you could bill between between 120 and 180 per hour likely.

4. Get your prime contractor to sponsor your company for a facility clearance.

5. As vacancies on the contract come up, hire people in your network. On a straight time and material contract you can probably make around $40k-$60k per person.

6. When you have a big enough company you can bid on smaller contracts and become a prime.

If you’re interested I wrote a book on the first phase of this plan (becoming an independent contractor). Currently have 6 employees and a business partner. Working on growing it so I don’t have to billable work anymore.

1099fedhub.com

I think the easier route has changed now. This advice was definitely true 10-15 years back, but I think now the entire sector is open to startups.

I would advise that startups team up with a larger player who can actually win contracts as a prime, and push out product together with them. It's sort of the middle ground between contracting and going alone. So I would suggest starting straight from #2 above up to #5. The biggest con is that you're completely dependent on the big company for survival, so you'll have to make them equally enamoured and dependent on your tech/product.

Just curious. "Defense" in this article and in "DoD" is a euphemism for military, isn't it?

I mean: these technologies (and the DoD) are perfectly capable of offense (attack). All while they prefer to call themselves "defense". This is just PR-speak right? A way to look good in the eyes of those who do not understand new speak?

You’re 76 years late to this party :-)

(That’s when the US Department of War was renamed to the Department of Defense)

It was “National Military Establishment” from 1947-1949.

According to wikipedia, the new organizational structure was created in 1947. In 1949 it was renamed to “Department of Defense”.

(I had just looked this up for my sibling response to parent)

The US Department of Defense has had that name since 1949. From 1789 to 1947, we (USA) had a Department of War. “Defense” probably qualified as “new speak” at the time, but after nearly 80 years almost everyone understands what it means.

It’s also when we shifted to permanently having a large standing army. The generous reading of those events is that by the end of WWII it was clear that a “great power” wouldn’t have time to re-arm for the next big war, because the pace of war had increased so much and maneuver combined-arms warfare was both very effective and required far more training than older approaches, and “defense” reflects this shift to maintaining a large military apparatus even in peacetime.

Less-generous readings of the events are obvious and common :-)

The DoD was named in 1949, after the end of an awful world war, and there were a lot of moves made to make sure something like that never happened again. Changing the name won't prevent war but maybe it'll remind some people why they show up to work. It's better to name things what they should be, rather than what they are criticized as being at their worst. Don't underestimate nominative determinism.

> It's better to name things what they should be, rather than what they are criticized as being at their worst.

This is known as propaganda. The US has done nothing but go on the offense post-WWII. There was no “defense” in Korea, Vietnam, Iraq, Afghanistan… those are all examples of US aggression and imperialism.

There was no "defense" in Korea? There at least was defense of Korea.

Department of other's democracy's defenses, because freedom. Murica.

I think you’re missing my point.

If you don’t think their actions back up their name, would you rather them work towards changing that, or would you rather them change their name?

I’d rather they were completely defunded and by having a name that matches their actions that is more likely to get support. There is a reason the military goes to great lengths to whitewash their actions, they require at least nominal support from the citizens funding them.

I don’t think that would do what you hope it would. Undermining the US military would not result in the US not having a military. It would result in the US having a different military.

> work towards changing that

But if they demonstratably didnt.. Then it's propaganda right?

And if they demonstrably did?: https://en.wikipedia.org/wiki/Charter_of_the_United_Nations

Yes

“The best defense is a good offense”

It doesn’t mention anything about the limited talent pool you’ll be pulling from to build war tech. I for example would never use my skills to enable or empower the military and I’m sure I’m not alone. As an industry we should resist building weapons and spy tech, even if it pays.

What are the most morally acceptable industries on offer that we should use our skills for instead?

Anything that doesn’t involve killing humans and destroying the environment?

I was hoping for a few specific examples. You might be able to find something with more abstractions between yourself and the killing and destruction, but if there are any 'pure' industries out there, there certainly won't be enough jobs for all of us.

Mass data collection, recommender systems, social networks and AI models built on "fair use" of everyone's hard work of course.

They're all such noble pursuits, so much better than defense! Just look at how much they've done for the world.

It's never not funny when people in our industry, especially those working at giants like Amazon, Google and Microsoft, whine against defense projects claiming to want to do good for the world while being responsible for so much harm. Shows that they're more than happy causing any amount of suffering as long as they don't have to see it themselves in the form of dead bodies and ruined cities.

Otoh the superficial social justice warrior types self-disqualify so you don't have to put up with their politics. There are some reasonable objections - say if your job was to make a chemical weapon more deadly - but lots of good military tech that keeps people safer. If someone wants to paint it all with the same brush and manipulate people into clicking ads or algorithmically denying insurance or whatever other tech thing instead, I'd be happy not to work with then.

Please don't post ideological battle comments. It's not what this site is for, and destroys what it is for.

https://news.ycombinator.com/newsguidelines.html

Yeah, well, you know that's just like, uh, your opinion, man. And not a very intelligent one. Ask Ukraine.

It is my opinion as I have a functional moral framework and the ability to think for myself instead of blindly digesting government propaganda.

Please stop posting in the flamewar style. It's not what this site is for, and destroys what it is for.

https://news.ycombinator.com/newsguidelines.html

Reasons not to build a "defence" startup: your work will directly or indirectly be used to kill humans.

How do you prevent yourself from being nuked ("just be nice" won't cut it)? I think mutually assured destruction has proven its merit at this point.

Iraq and Afghanistan weren't about to nuke the US. I don't see the moral justification for work going into those wars, even if there is a possibility that it might not be used solely for invasions, imperialism and bombing weddings.

We are talking about defense in this thread, not offense, but you are in a way proving my point. If Ukraine had nukes would Russia have invaded? Would the US have invaded Iraq if it had nukes?

What? The defense industry is not just for actual defense though, it's just a euphemism for military industry. Am I misunderstanding something?

Just gonna point out a few options to prevent future invasions like Iraq and Afghanistan:

1) Elect politicians who will not invade Iraq and Afghanistan.

2) Don't have a military and hope you're not the next Iraq or Afghanistan.

that's not entirely true. Did funds lots of very interesting research that enhances quality of life for citizens.

But yes an absurd amount of capital gets spent on, well, defense!

I'm sure the various autocratic, dystopian states around the world that would love to see functioning democracies be smouldering piles of rubble are hoping that we think this way...because they certainly aren't thinking that way.

Reply on news.ycombinator.com