Capitalism 101 – The Rising Tide of Bullshit

“A rising tide lifts all boats.”

I’ve got no idea how many times business people and venture capitalists have said that to me, but if I could, I would have personally, slowly, stabbed them in the heart for doing so.

Here’s why: it’s the worst and most transparently dishonest way to try to excuse inequality. There is another variant of the same saying, which is: “Which would you rather have, a big slice of a small pie, or a small slice of a huge pie?”

Low tide in the Bay of Fundy

If you hope to ever actually get money or work with a person who says that to you, you are usually the one who is on the short end of the equality stick (or they wouldn’t be saying it) so you have to suck it up and knuckle your forehead and mutter, “yes your lordship.”

The correct answer, which you can only give if you really don’t care (i.e.: the pitch meeting is going horribly wrong) is: “I want your slice of the pie.”

Let’s look at some typical context: you and I, dear reader, are the founder team for a start-up. One of us is the CEO (“businessperson”) and the other is the CTO (“idea person”) and we have an idea that we think is going to be the next Facebook only not a disgusting time-wasting sewer of advertisement content. We’re trying to raise $5mn so we can build this business and we’re sitting in one of the many art-filled meeting rooms of Andreesen Horowitz’ $17mn office building on Sand Hill Rd in Palo Alto. [Marc Andreesen wrote a shitty browser, once, now he’s a billionaire; Andreesen Horowitz is his venture fund] Since $5mn is a tiny deal for a top-tier VC we’re meeting with an underling. We’re the 5th meeting already, for the underling. We’re on our 3rd meeting today. So the underling says some stock bullshit about how Andreesen Horowitz likes to “be in early in the cycle” when they really can be involved in shaping the company. That’s stock bullshit: what he’s really saying is that they want to invest in a controlling interest in any start-up, or forget it. They want to buy 51% for that $5mn. And you say, “sure, but we’re not interested in taking big money until this thing proves itself out.” That’s a polite way of saying “ha ha ha yeah I bet, but that would value us at $10m for the whole company and we’re going to be the next Facebook. We’re worth more than that.” Then the underling smiles and says “a rising tide floats all boats!” and you leap across the table drawing your katana with feral smoothness and…

Andreesen Horowitz, what a dump. I’ve been there for at least 4 pitch meetings.

No, not on the hand-rubbed hickory flooring. That would be wrong.

You’ve got the picture. What’s going on? “We are powerful and can set the price of your stock because we have all the money and you don’t” is the primary subtext in all of that. The secondary subtext is the idea of the rising tide: basically “sure you are giving up a huge chunk of equity, but if this thing really is the next Facebook, those shares are going to be worth billions and it’s silly to walk away from a deal over a measly $5mn right now. Of course the problem is that the billions are aspirational and the $5mn is very real. The slice of the pie argument is the same thing: if you stay small and lean and hungry and don’t have big money behind you, you may never amount to much; this this could be the next Facebook (huge pie) or it might be an also-ran (small pie, perhaps a cookie) and if you’ve got 50% of a multi-billion dollar pie that’s much more than 50% of a $30mn pie. That’s all true, except – again – all of this is aspirational. That’s why it’s a huge violation of the rules to not pretend to play along. It will end the meeting very, very quickly and on a bad note.

Venture capitalist assholes want to convince themselves that they offer something other than just access to money. That’s largely a lie, since the 90s, when the SEC started cracking down on some of the more obvious ways that VCs were dealing high cards into their own pockets. Nowadays VCs mostly want to buy low, sell high – buy a chunk of a start-up that’s the next Facebook for next to nothing, then sell it for hundreds of millions of dollars. “Rising tide!” What they’re really saying is “we want half of the functional space on your boat so we can have occasional deck-parties while you guys work the ship. And, yes, a rising tide will raise us all.” Another quip that has come to mind is “a rising tide also lifts the barnacles, you fucking venture capitalist parasite.” I don’t see any reason to be mean to the poor barnacles; they are just following the path nature laid out for them. The “rising tide” scenario means that – because they have money and you do not – they get to position themselves to grab an obscene amount of the final pie if it’s worth a lot. And, in the meantime, they don’t actually do a fucking thing to earn it; you and your crew will be doing the work.

In other words, it’s the same shit where your boss wants to not only put the boots to you economically, they want you to love them for it.

Which brings me to Bill Gates and Elizabeth Warren. [nyt] Gates said:

I do think if you tax too much you do risk the capital formation, innovation, U.S. as the desirable place to do innovative companies – I do think you risk that.

Considering that Gates’ billions come from his founding one of the most derivative, least innovative, software companies ever – one which (arguably) stole the design of its MS-DOS operating system from IBM, and its windowing system from Xerox PARC – it’s a bit ironic to see Gates worried about innovation. But I assume he’s talking as a venture capitalist would, now, not a technologist. So: you risk the ‘capital formation’ and ‘innovation’ – what he’s talking about, there, specifically, is that the stream of innovators and founders would stop crawling to venture capitalists’ palaces to kiss their rings, because of the threat of paying more taxes. See this guy Gates is held up as a business genius because he’s a billionaire, but my horse Otto is smarter than he is – Otto would say, “but Bill if you’re worried about innovation why don’t you just invest in tons more start-ups and settle for a slightly smaller scoop of the oats? After all, a small scoop of a healthy sized sack of oats is better than no scoop of oats at all, amirite?”

Why don’t any of the business press give Gates back Otto’s response? Could it be that they’re just ignorant corporate flacks? Or maybe they’ve got less sense than a horse who is not legendary for his intelligence, even among horses. That’s the simple, snappy response. There’s a better and more sophisticated one, which Elizabeth Warren (remember: she’s a “capitalist”) would not utter, namely: “But Bill, the entire system under which we finance innovation in this country amounts to a tax prior to earnings. Every VC who takes equity in a start-up is ‘taxing’ it if they ever succeed. Innovators and founders are perfectly willing to play the game under those circumstances; why ever do you think they’d walk away from the game en masse if there was one more hand in their pocket grabbing a little slice of their profits?”

Let’s do the fictional math: we raise our $5mn and our company takes off like a rocket. 5 years later, it’s worth $3bn and we’re bought for cash by Microsoft. Our VCs get their $1.5bn. The founders and staff divide up their $1.5bn. Let’s say each of the founders’ shares is worth $500mn. Since we’ve been holding the shares for 5 years it’s long-term capital gains and the tax-rate is fixed a maximum of 20%. So I owe $100mn and – oh, fuck, did you just see the razor blade in the scoop of oats? I’m not a billionaire! I write my tax check for $100mn and I’ve got $400mn left over and I’m not a billionaire so I totally do not give a fuck. I will, however, own 100% of the next company I start, since it will be self-funded, and the venture capitalists will never say “A rising tide…” to me, again. In this scenario, the only person who is “dissuaded” from anything is the venture capitalists, who are unhappy that they will have to pay an elevated tax on the $1.5bn that they got for doing exactly jack shit; i.e.: lending the founders of a start-up $5m. Gates is trying to say that him having to pay a bit more to help foster innovation means he’d rather take his marbles and go home. That’s fine, Billo – you just priced yourself out of that market, but there will be venture capitalists who would be perfectly happy to see things the way Otto does: some oats are better than no oats; we have a deal.

And Elizabeth Warren, capitalist tool, pretended to take Gates’ bogus argument seriously instead of saying, “well now that you mention it, if I am elected we will start a program that would replace the venture capital system as a way of fostering innovative start-ups, especially ones run by women and minorities. So you won’t have to worry about that; what you’ll have to worry about is hungry well-funded competitors. And I hope they eat your lunch.”

------ divider ------

Re: self-dealing venture capitalists. In the 90s, VCs realized that they could manage the companies they owned stakes in as part of a “portfolio” of companies, and have them do business with eachother. That means, basically, that their sales get artificially inflated. I remember when Akamai started up; their entire network fabric was built atop gear from Juniper – instead of the usual choice for networking gear, which would be Cisco. Why? Their VCs also owned a chunk of Juniper. And they told Juniper to use Akamai. A guy I’ve known for a long time was one of the senior network engineers at Akamai and I remember him complaining that they had crates full of Juniper gear that they had bought and warehoused – millions of dollars worth – then they ran their production network on Cisco gear. The VCs didn’t care: both companies’ bottom lines had been artificially boosted by millions of dollars. This was back in a time when a couple multi-million-dollar deals were all you needed to go public and have a market valuation of $billions. Of course, the market(s) caught on to that, which had a fair bit to do with the tech stock crash; suddenly everyone realized they were overpaying hugely for companies that were basically shit – but the venture capitalists had already sold their shares and gotten out and were ready to move on to new markets.

In case some of you are thinking “wow, Marcus sounds bitter.” No, I’m angry, not bitter. I’ve done great. Many of my associates over the years are now fabulously wealthy; I actually know billionaires. Some of the start-ups I’ve been involved in went public or got bought and I’ve done well enough that I was able to step off the treadmill and not have to go around sniffing venture capitalists’ asses. If I wanted to try to make gigantic amounts of wealth, I have enough, now, that I could fund my own start-up, which would mean that I could die with more money that I could possibly do anything with, or lose it all. Instead, I chose to be a blacksmith. Who could be bitter about that?

Another answer to Gates would be, “Sod that, Billy. I’d cheat on my taxes just like you did.”


  1. secmilchap says

    Brigand Bill’s success had two more big scores to add to the two purloined TechBits: He appears to have violated anti-trust law by forcing all hardware vendors to make his OpSys boot automatically upon startup of ALL their products, and, with tRumpian dance steps, kept the Feds from digging into his unethical behavior until he’d accumulated enough cash to buy legislators and others. Can’t fault him for cunning, despite his claims to “innovation”.

  2. says

    The correct answer, which you can only give if you really don’t care (i.e.: the pitch meeting is going horribly wrong) is: “I want your slice of the pie.”

    I’d answer differently:
    “A rising tide lifts all boats.”
    “Oh, in that case I assume you wouldn’t mind if the two of us swapped our boats? I want your boat instead.”

    Andreesen Horowitz, what a dump. I’ve been there for at least 4 pitch meetings.

    Looking at this photo, my first thought was: “How do I sell my artworks to such dumps?” I need money, and my art is better than what’s framed on the walls there.

  3. Owlmirror says

    Marc Andreesen wrote a shitty browser, once

    Was there anything computer-related from that era that wasn’t shitty?

  4. Andrew Dalke says

    “(arguably) stole the design of its MS-DOS operating system from IBM” — I think you mean Digital Research, which developed CP/M.

    Owlmirror @#5: Doom, more specifically the Doom engine, came out the same year as Mosaic, and isn’t in the shitty category.

  5. cartomancer says

    So, if you’re the CEO and I’m the CTO in this little enterprise we’ve cooked up, the important question is: have you stiffed me out of my share of the profits and sold me down the river yet?

  6. komarov says

    If ship analogies are insisted upon, I’d just like to point out that the lower decks always flood first and the lifeboats are topside. I’d imagine below is also where you get the rust, the damp and the bilge rats. Well, the latter may have migrated to the top of the USS Venture (Capitalist).

  7. Ridana says

    Not being a sailor, what I always envision when I hear the floating boats metaphor is little boats chained to the dock, where they can only rise to the level the short chains allow, and then get overtopped and sink, while the yachts sail merrily away on the high seas.

  8. says

    John Morales@#9:
    and I’m rather surprised Marcus didn’t know the history.

    MS-DOS and its history are fairly uninteresting to me.

    My recollection was the IBM had contracted someone to design the operating system for the PC, but that Gates came in and bumped them out by buying the rights to the IBM O/S and substituting Microsoft’s. That’s pretty close to what happened, but I got wrong who Gates ripped off. In retrospect, a small time player like Microsoft would not have survived ripping off IBM.

    Microsoft pulled a similar trick on Spyglass, which is how Internet Explorer came to be. Microsoft went to Spyglass, who had an excellent and fairly fast browser (Microsoft appeared to be unable to write their own) and negotiated a deal whereby they would pay Spyglass a certain amount for every copy that they sold. When they inked that deal, Microsoft immediately began giving the browser away, so there was no copies sold and no revenue for Spyglass, who went under fairly quickly. Another example of Microsoft’s innovation.

  9. says

    Andrew Dalke@#6:
    I think you mean Digital Research, which developed CP/M.

    I mis-remembered the relationship between IBM and Digital research; I thought that IBM was working on their own O/S when they were paying Digital Research to do it, instead.

    CP/M wasn’t bad. I ran it briefly, in a version called ZCPR/3. It took a while for me to realize how limited PCs were, at which point I turned my PC into a games-only machine and got an Amiga, which at least had a decent C compiler and a real O/S that wasn’t just a jumped-up program loader.

    Anyhow, yes, I made a mistake about who Gates ripped off.

  10. dangerousbeans says

    “Wouldn’t you draw your Katana with feline smoothness?”
    How about we compromise and go with feral feline smoothness? Some of the local feral cats are pretty fucking scary

    Does all of this venture capital, stockmarket finance BS strike anyone else as just creative accounting and math tricks? It strikes me as decoupled from anything actually real

  11. says

    Was there anything computer-related from that era that wasn’t shitty?

    Sun Microsystems was producing systems and an operating system that were really good. Unfortunately, they were just about to begin to collapse under their own weight and McNeely spent a huge amount of money going after Microsoft’s monopolistic strategy. [I wish someone would write a definitive history of that time; that whole case was fascinating – Oracle and Sun writing the legal briefs for the DOJ, which had been claiming they could not afford to take on Microsoft…]

    The aforementioned Spyglass browser was exactly what a browser should be, at that time, and was solid and fast and simple. It was a brief stand-out before the browser wars started and everyone threw as many features at browsers as they could think up.

    DOOM was great; really amazing given the machine resources it had at the time.

    Another great bit of computing that was coming out at the time was QNX. An O/S environment that could produce a single system image out of multiple separate systems, with process migration, hard real-time, user-mode filesystems and a networked filesystem… It was great stuff. They released a single floppy image that booted the full O/S, window system and a browser, and it was ridiculously fast and stable. Unfortunately, QNX just started to get cooking in the market when free UNIXes and Linux came and sucked all the oxygen out of that niche. I believe QNX is still the kernel of Wind River’s real-time OS.

    Also around that time, UUnet invented the “internet service provider” as we know it today. That was pretty cool, too.

    And then there was BSDi – bankrupted in a spurious lawsuit by AT&T, the company failed and threw its intellectual property out on the ‘net; Linux and all the BSDs would not have happened if not for BSDi.

    There was some good stuff out there. It’s not all shit.

    I remember, however, the moment the world changed. I was in the data center at Sterling Software in Dallas the day that Netscape went public. And we all watched the stock scream up and up. Kent L, who I was working with at the time, said that it meant that computer programmers were going to be the new rock ‘n roll stars. I thought it was shocking that a company could be valued so highly for 1) giving away its product and 2) the product was permanently in beta-test. In case you’re wondering, Netscape is now owned by Verizon, who got the shattered remains along with AOL.

    There was also the Mozilla project, which tried to give away truly excellent free software and discovered that unless you have a revenue stream, it doesn’t matter. Google tweaked their business model slightly, and cleaned up.

  12. says

    So, if you’re the CEO and I’m the CTO in this little enterprise we’ve cooked up, the important question is: have you stiffed me out of my share of the profits and sold me down the river yet?

    Did you just notice that ice-pick like thing sticking out of your back? Is that what you’re referring to?

  13. cvoinescu says

    Marcus, I very much enjoy your “Capitalism is a disease” posts. I’m sure the world will not cease to provide you with opportunities to write more of them.

    CP/M was okay (jumped-up program loader is perhaps too harsh; it also had a single-directory filesystem, complete with all-caps 8.3 filenames). I got to use it a little in high school. Our CP/M machine had an 8080, and its terminal was a VT52 clone with a Z80, clocked faster than the 8080. However, there was a PDP-11M clone in the same room, so there was no contest.

  14. brucegee1962 says

    @17 cvoinescu

    I still have files with all-caps names in my directory, copied from computer to computer to computer all the way back to old CP/M floppies.

  15. colinday says

    @Marcus Ranum

    And then there was BSDi – bankrupted in a spurious lawsuit by AT&T, the company failed and threw its intellectual property out on the ‘net; Linux and all the BSDs would not have happened if not for BSDi.

    How did Linux depend on BSDi? The latter was not released until January 1992, while the former was released (0.0.1) in August 1991.

    Also, BSDi settles with Novell/USL in 1994 and lasted until the early 2000s.

  16. says

    @colinday: berkeley released the net2 tape, which was a mass of code considered “AT&T free” in order to help foster non-AT&T UNIX versions. BSDi was one of those, gnu/unix another. Net2 included big chunks of BSD including the IP stack and the sockets kernel interface as well as user-land utilities like a lot of stuff in /usr/bin and /usr/ucb. BSDi was funded by USENIX and UUnet to fill in the missing pieces to produce an AT&T free BSD.

    BSDi lasted a long time because they did some good licenses on embedded systems versions of BSD. In fact the early versions of my TIS Gauntlet firewall ran on BSDi – we sold tons of it.

    I haven’t talked to the BSDi folks since the mid 90s but at that time they were under a huge financial and legal shadow – it’s hard to bet your tech strategy on a company that’s being attacked by a giant company like AT&T. A lot of companies started using FreeBSD or OpenBSD (my NFR product was built on a highly customized BSD kernel – initially OpenBSD later FreeBSD for the multiprocessor support) because it was stable by then, and free. None of that would have happened if it were not for plucky little BSDi who got hit by a train named AT&T.

    The /usr stuff in BSD net2 wound up getting folded into gnu, which jumped on the Linux bandwagon and Linux went its own way with its own IP stack, and a lot of changes happened in /usr land.

    Michael Meissner (If I have that right) is also a hero of that story, initially developing a replacement C compiler – gcc – that quickly superceeded AT&T’s cc. The C compiler, developed at Bell Labs, was a critical part of the UNIX environment and gcc was a great blow for freedom.

  17. says

    Novell/USL was a weird story. I forget how they got spun off and funded but they tried repeatedly to claim ownership over various bits of UNIX and Linux. There was the famous suit against IBM in which they claimed that IBM’s mainframe Linux was basically stolen code. It was not, because it had been specifically re-written to be AT&T code free and the implementation and designs were different. There was a lot of overlap – you implement the hardware specific parts of virtual memory the same way because the hardware demands it. But the IP stack in USL (which was SystemVr3 I think) was completely different, for example – BSD used sockets/select and SVR3 used shared memory segments/spinlocks.

    I was a non-testifying expert on that case, specifically decoding and comparing some pieces of kernel implementation between the two systems. I quit when I realized what USL was trying to do and also that they had no case. I tried to point that out in a concall with legal and the lawyer started screaming abuse into the phone, so I hung up and presumably someone else re-did my work.

    It seems like it was all another lifetime ago.

  18. says

    jumped-up program loader is perhaps too harsh

    That was a reference to MSDOS, which basically was that. Sorry if I was unclear.

    I forget how DOS did it but basically when you started something running, most of DOS went away except for a little virus-like bit that hung around and came back when the program exited.

  19. John Morales says

    Was better than that, with TSRs. I wrote one for screendumping my Hercules graphics card, back in the day; and I also remember Sidekick as a great utility which put Borland on the map.

  20. Andrew Dalke says

    All histories I know of have Stallman as the main force behind developing gcc.

    Michael Meissner is a long-time GCC developer. His home page at shows his early employment at Data General, developing “GNU Compiler for Motorola 8800 and Data General compilers for MV/Eclipse, Eclipse, and Nova computers”. quotes a 1987 Usenet post by Meissner, then at Data General.

    Data General was one of the first vendors to ship gcc as the default C compiler (see ). Gcc 0.9 (from 1987) supports only the Motorola 68000 and Vax chips, not the Motorola 8800. The GCC source contains many copyright statements by Stallman, and none by Meissner or Data General.

    That is, Meissner seems to be an early GCC user and contributor, but not one of the original developers.

  21. khms says

    I think this needs more history correction. All of the following is from memory:

    First, DR developed CP/M. Then, IBM developed the PC (basically a rip-off of the Apple II+ with a newer (but not faster) CPU and more RAM; the chief designer had one at home). They needed an OS, DR had one, so they went to talk to DR. DR’s boss, however, did not realize the importance of the meeting and essentially blew it off to have fun with his plane instead. So IBM still needed an OS. MS got wind of that and wanted to get in, but they didn’t have an OS. But there was this guy who had written a clone of CP/M for the 808x (“Quick and Dirty Disk Operating System”), and they could get that. So they were set. Putting an MS Basic in ROM on the PC was another parallel to the Apple II+.
    Oh, and getting their GUI idea from a PARC visit was Apple with the Lisa and then the Mac – unfortunately, they did NOT also get the idea of object-oriented programming, that came much later. At the time, there were rumors MS got the idea from Apple.

  22. says

    Capital formation is a genuinely useful function, but since it is by definition creating large piles of cash it attracts grift like nothing else. Arguably this function should be government run.

    Andreesen-Horowitz is, even by SV standards, a bunch of fuckwits. They seem to have pioneered the “well, we invested in it, so its valuation must be enormous” model of inflating company values, a system which is serving Softbank hilariously poorly these days. VCs used to be better, but they’re just trash now, I’m pretty sure. I’m by no means an insider, but I have kept careful notes from various brushes with various insiders over the years.

  23. says

    As for computer history, it is a mistake to try to trace specific ideas to specific places.

    The IBM PC arises after a long period of innovation, as various and sundry hobbyists and corporations attempt to crack the code on a computer that can be sold to home users. Apple and Radio Shack have managed some pieces of the puzzle on the mass-market side, but the S-100 community has been building stuff and developing software for these systems for quite a while as well. Sinclair is selling incredibly cheap and weird computers. TI has their even weirder 99/4 and 99/4A systems. There are probably a dozen lesser systems I can’t recall or never heard of.

    There is a stew of ideas in play, and everyone is stealing from everybody else. BASIC in ROM was just a thing. Every system I named above had BASIC in ROM. Nobody has any “operating system” for any of these things. You don’t even need a Disk Operating System when you’re mainly loading programs from either a cassette tape, or a ROM cartridge.

    The point of these computers was to “run a program and later, run a different program” the idea of an “operating system” was absurd. You had device drivers which were, often as not, implemented as ROMs physically adjacent to the devices, which provided simple APIs that programs could use to access the device. Some management layer to mediate access to things and allocate resources would have been a nonsensical waste of resources.

    The triumph of Windows, and at the other end of the spectrum IBM’s z System, is the extent to which they were able to smoothly manage the transition from every program runs alone on bare metal into a fully managed operating environment. The IBM guys realized the power of virtualization far earlier, and did a better job.

  24. says

    Related more to the post title than anything else, it may amuse you to know that the FT’s Alphaville blog (so, not quite the FT proper, but not a million miles away from it either) has an ongoing series entitled “The Entire Economy is Fyre Festival“.

    That sounds great; I’m on it!

    It’s amazing how whiny a bunch of rich people are that their stupid festival was a ripoff and the champagne wasn’t cold enough, etc. Bunch of snowflakes. I mean, sure it was a horrible mis-managed ripoff but…

  25. says

    Andrew Dalke@#25:
    All histories I know of have Stallman as the main force behind developing gcc.

    I clearly mis-remembered. Stallman was nowhere to be seen at the USENIX C++ conference, but that probably had more to do with gnu/USENIX politics than anything else. During that period, I was one of the people who just kept his head down and did stuff, avoiding the various movements. There was a lot of blatheration going on. I was mainly amazed at how stupid Sun was to unbundle their compiler suite, which was what gave such an impetus to gcc that it rapidly became tolerable and eventually we started switching.

    I admit that I’ve sort of edited Stallman out of my memory. Anyhow, let me stand corrected.

Leave a Reply