Risks and opportunity costs

The context of this post is about an important decision, at which company I am going to work, that I am finalizing now.  Every single decision you make between such scenarios carries a risk.  You can't eliminate risk.  You can, to some degree mitigate it by chosing carefully.

The context is that the two organizations are, in their own ways, fantastic.  One is a startup in storage, one is an established player in semiconductors.  The startup has some, quite frankly, amazing plans going forward.  Its been presenting its vision, and it aligns very well ... exceedingly well ... with what I was thinking about back in the halcyon days of my own company a decade ago.  They are doing financially well.  I can't share what's been shared with me, other than to say I'm not worried about them running out of money or runway.  They could IPO eventually, but my best guess for an exit is an acquisition by a partner looking to add to their portfolio.

The risk is, of course, that they never hit escape velocity: can't IPO, or aren't acquired.  This is the common reality of startups in this market.  They are expanding their focus beyond storage into sensible adjacent areas usually served by software/SaaS type shops, and showing ways to add value to those shops.  Very early days for this work, and its unknown how this co-opetition will be received in the market.  Hence more risk.  

But there is no earned reward without risk taken.

The other, established firm is in the semiconductor space. They make CPUs and GPUs.  They are not the dominant GPU vendor for AI, but are working on catching up.  They have a stack that I've personally used before (along side the dominant AI and accelerated computing stack).  And if you've been a reader of this blog since the beginning, almost 20 years ago, you'd know that the first iteration of my own startup was as an accelerated computing company.  Designing and building accelerators for offloading matrix/vector computing.  Investors, 20 years ago, were reluctant to put money in, as they didn't see value in accelerated computing, or the potential size of the market.

So we pivoted after a few years of failing to raise that capital we needed.   I still have our pitch decks somewhere, I may for laughs, put them up here for posterity.  

The dominant GPU vendor has done an absolutely killer job at executing something quite like our original 2003-2005 based business plan.  They are worth, currently, north of $1T USD.  

Yes, this is one of many reasons why I say the vast overwhelming majority of venture capital is dumb money.  Success they have may be correctly attributable to luck.  Probably not a popular view in some circles, but hey, its my opinion, and backed up thus far by the empirical evidence I see.

For those who don't recall, my startup was taken out behind the barn by our bank, and shot in the head.  Assets sold, IP dispersed.  

I swore I'd be very careful about considering startups going forward for employment.  I lived the risks.  I know the ups and downs.   I've dealt with the personal financial destruction.

Similarly, for challengers to market leaders, you have to have nearly perfect execution, and very strong focus.  You cannot chase everything, you need to pick the battles you can win, and win.  Those battles have to be meaningful for the larger market, something that users will care about.

As I've always told people, you need to target ubiquity in the accelerator market. Which means you need to make your platform work in the widest possible manner.  You need to trivially enable people to develop on low cost/low power systems, and make it easy to move to higher power systems later.  Usually not large code changes for this, the initial "port" is what matters.

The semiconductor company hasn't done a great job of targeting ubiquity.  This is unfortunate.  If I joined them, it would be in a customer support role, so my ability to help drive the sort of change that would be needed would be, at best, quite limited.  I'd get to work on the accelerator side, and I can do the job they want.  That's not a problem.  That is, the risk isn't in the position, its in the company execution, and investment side.  

Have they, or will they, invest in enough resources to massively increase their accelerator TAM.  They have excellent design wins, the linux laptop I'm using now, is using their technology.  And that of their competitor.  But they seem to have relegated an otherwise functional, albeit 3 year old technology, to being unsupported.  Which means I can't use their tech here.

I can however, use their competitors tech.  On the same platform.  This is what I've been using for my own experimentation.

This isn't risk, this is opportunity cost.  The alternative decision, not to scrap support for non-obsolete hardware, would likely be paying dividends now.  Many millions of client machines, capable of being used for code development and testing, can't be used for testing now.

The question I have, and I've posed, is whether or not there is a recognition of this within the company.  Of course, this goes hand in hand with where one invests strategically.  Where do you put your (finite) money?  It is more finite than their competitor.  So they need to be more careful.  

The startup doesn't quite have the same opportunity cost issues.  The technology they are deploying is very smart, and compatible with most of their users infrastructure.  Their business plans make a great deal of sense to me.  They don't have to force particular kernel versions for clients, or use proprietary tech outside of their platform, to enable others to interact with it.

The role there would employ my technical and soft skills.  I would naturally fit into the organization, and can clearly see their challenges, and how I could help.  I could make, I believe, a strong contribution to their success.  Their current mission is very well aligned to my old companies mission.  Its scary how well aligned it is.  I would enjoy working for these folks.

That said, the others in this space aren't standing still, and have been in market longer.  Some are public companies, though all have valuations not large factors off the current valuation (by investors) of the startup.  The risk here comes from any liquidation preferences investors have.  Remember, while VCs are generally dumb money, they often negotiate investments in such a manner as to make sure the money flows to them first, and sometimes, only.  So if the startup offered N options, and at IPO or acquisition, all the capital was absorbed by the VCs, the risk is that the value of those N options would be $0 USD to me.

This is why startups are intrinsically lottery tickets.  And hope is not a strategic plan.

The realities I face.  Scalable Informatics death wiped me out financially.  At age 51 I had no money saved for my daughters college, or retirement.  All of it was in the company.  Crawling back out of that hole has required quite a bit of sacrifice.  We somehow managed to be able to pay for her education so she wouldn't have a huge amount of student loans.  And put money into 401k, while redoing the house mortgage to pay off balances and debt (including when the *&(%%$%^ bank sued us).

Today, I'm looking at how to complete house repairs (been doing that for a while now, and have basically 2 things left), while saving away for retirement.  As I am assuming that the social security system that I've been paying into for nearly 40 years is going to be bankrupt before my retirement, my planning cannot include that being available.

The issue is that I am closer to 60 than 50, and my decision has to be based upon what is best for my family in the long run.  Both offers are good, though one is somewhat better, from the financial basis that matters.  The one that focuses on my family.

Risk exists for both.  Semiconductor company has legacy bits which aren't very additive to margin.  Hence that missing money can't be plowed back into R&D.  The startup could (though not likely) miss revenue targets, or need more capital to gain runway, which means dilution and lower valuation than hoped for.  Not to mention the liquidation preferences.

I've been thinking this through for a day or so.  I have my decision, and I think I will be signing papers this day.

You can't escape risk.  You don't live/work in an isolated bubble, you often need to be cognizant of your decisions impact upon others.  

But I do like underdogs.  I really do.

Show Comments