Going, going, gone.

Software Will Replace You

Going, going, gone.It doesn’t matter who you are – software is going to take over more and more of what you do.

Every once in a while, someone writes something that gives voice to an idea you’ve struggled to articulate. Marc Andreessen just did that for me with Why Software is Eating the World.

The essence of his argument is this: software-based businesses are rapidly displacing businesses that aren’t software-based. Amazon is replacing Barnes and Noble. Zynga is replacing Milton Bradley. Netflix has already replaced Blockbuster.

Business Process Automation

This isn’t just happening at a macro, industry level. It’s happening at the micro level too – right down to the individual tasks you perform at work. Just look at the field of business process automation, a class of technology that, until recently, was only seen in very large companies. That’s changing now as companies like Salesforce simplify and reduce the cost of capturing what we do in software.

Capturing “business processes” isn’t really anything new. Management gurus have long talked about documenting “best practices” and we have a long history of industrial pioneers like Henry Ford who have obsessed over streamlining the way we organize ourselves to get work done.

When we capture those good ideas and best practices in software though, something important happens. Those processes can now be executed in software, and that enables us to automate them and scale them up in ways that we simply couldn’t before.

This ability to virtualize work is showing up in lots of places, and with varying levels of continued reliance on humans. Manufacturers have virtualized large portions of their design processes with Computer Aided Design, but still need talented human designers to lead these processes. On the other hand, Automated Teller Machines (ATMs) disperse cash for us without any real assistance from humans, aside from periodic restocking. That’s a process human beings used to do for us not long ago, but it was one that was simple enough that banks and customers agreed that software did a better job, and for less money.

And that’s usually the catalyst for software virtualization: lower cost, better service.

Engagement Software

As software gets more sophisticated and more widely integrated into the nooks and crannies of our organizations, the range of business processes that we can virtualize will steadily increase.

One area where we’ve seen huge jumps in process virtualization are the communication and engagement processes that organizations use to connect with the external world. I see these processes as a kind of  “organizational membrane” and one of the greatest sources of competitive advantage these days stems from using software to scale up these connections at lower cost and with better service. That’s how Amazon provides a ‘personal’ shopping experience to millions of customers everyday. It’s how Netflix, Zynga, and Facebook serve their huge customer bases as well.

These firms dominate their respective markets because they know how to use software to virtualize the simpler, more straight forward aspects of connecting with their customers. It’s what enables them to scale so magnificently. As customers engage more deeply, and the level of complexity increases, the need for human touch becomes more important, so software doesn’t solve everything – and that’s the essence of the “engagement pyramid.” But what I didn’t address when I wrote that piece is that software is slowly working its way up that pyramid, replacing more and more of the way organizations engage their external constituents.

I Was Wrong. Humanity is Not Guaranteed.

I don’t like contradicting people – especially myself. But I have to correct something I wrote a few months back when I noted that “the organizational membrane may be augmented by technology, but it’s still fundamentally a human phenomenon.”  Thinking back on that statement today, I have to say that I’m not so sure.

Will we reach a day when software takes its current trajectory to a logical, if somewhat frightening, conclusion? Will software one day run our organizations? I’m not talking about “run our organizations” in the way our cars ‘run’ on gasoline. I mean “run our organizations” as in “Steve Jobs runs Apple.”

It may seem a bizzarre question, more worthy of a science fiction novel than a blog about people and organizations, but I’m quite serious. In our current economic system, where we run companies exclusively to maximize wealth for shareholders, what really would stop us from taking our organizations to this ultimate level of efficiency? Assume for a moment that it is possible, that Moore’s Law will eventually get us there. What would stop us from completely virtualizing our organizations by managing them entirely with software? What would stop us from completely stripping them of their human soul?

I’m not sure what the answer is. Are you?



  1. “What would (can) stop us from completely stripping them of their human soul?”

    Simple answer. We can.
    And we ought.

    We ought because we must.

    “In our current economic system, where we run companies exclusively to maximize wealth for shareholders…”

    How do we define our shareholders?
    How do we define wealth?

    How can we not see that we will slowly but surely self destruct as a civilization if we do not set limits based on our common good?
    Why would we make ourselves redundant?

    Exactly how dense are we?

  2. Kenny Braeckmans

    True! I could replace the management team where I work with a shellscript
    (the script would refresh facebook every 5sec, and use up bandwidth by going through random youtube vids)

    As long as it won’t replace getting shitfaced with my buddies, holidays, porn, a beautiful poem, etc… I’m fine with it, to say the least. Seems that tech will replace what we don’t want to do (kinda like what technology has been doing since… well, technology)

    Jobs are so 2010 anyway

  3. Software already runs many businesses. Consider the knowledge (included policy decisions) inherently embedded into ERP systems by their developers rather than by their nominal managers. Not to be too Luddite about it, but it seems to me that that the rise of automated management has concentrated wealth in the hands of those who own or control the production of that intelligence, rather than the people who are mandated to follow rationalized guidelines and processes. See Lang’s Metropolis for the full story.

    • Yep, on both counts. A big chunk of the value in IT systems is going to be bundled expertise. That’s the essence of the IBM consulting model.

      And yes, I think you’re right. I’d even say that it’s a more general rule that investments that make labor more productive shift wealth from the people doing the work to the people doing the investing. IT is one of the best examples of this.

  4. A singularity economy. A vision of progress that human beings imagined, but couldn’t fulfill. Their ironic evolution. A corporate matrix wherein humans play productive games. Each simulation increases efficiency and may be run by an artificial intelligence. Indeed it’s all quite sci-fi to think about, but one can’t help but wonder if we’ll merge as if processes within an algorithm. We augment our intelligence using an interface that fulfills the purpose of our collective soul. Now is that creepy, or inspiring?

    • Is it creepy or inspiring? That is the question I ask myself over and over.

      I can imagine a kind of benign software-based management that operated firms in ways that allowed them to co-exist with their surrounding far better than our current human-based management systems. I can also imagine software that was relentlessly and ruthlessly efficient in optimizing for profits.

      Today’s software is based on rules. Today we run the system and we have defined the rules that run that system. Could it be that we have a relatively small window in front of us right now to get these rules right? To set them up so that they don’t come back to haunt us in civilization-threatening ways down the road….

      Tomorrow’s software may no longer be based on explicit rules, but instead operate more along the kind of pattern-recognizing modes that our brain does. When that happens, it’s unlikely we will even understand what is happening within the software (anyone who’s ever done software testing can attest that we barely do even today). Few truly understand the trading algorithms on Wall Street today. But the system grinds ahead, optimized for one simple goal: profits.

      Looking beyond profits is hard. It’s nearly impossible to quantify how to truly live and let live. Nature has done it. But we seem to have struggled with it. Maybe we’ll just have to leave it for that which inherits our world to figure it out.

      Is that creepy, or inspiring? I just don’t know. But as a human, I think that chill I’m getting has more to do with creepy than inspiring….

  5. The negative side of this is that people are displaced, and unemployment expands. Finding work for lesser skilled people becomes a critical social factor. The positive side is that people are now liberated from the boring and routine to spend their time on the significant.

    As always, progress is disruptive: sometime frightening, sometimes exhilerating!

  6. How long until a bot with clever software can go into any home, find the kitchen and brew a pot of coffee? Monica Anderson gives us a partial answer in her presentation (next edge stuff)

    A New Direction In AI Research

    Monica Anderson proposes adding a new target to ongoing AI research efforts: We need to focus more of our attention on Understanding as opposed to Reasoning. Understanding requires using Model Free Methods. As a bonus towards the end, Ms. Anderson also speculates about the so-called AI singularity and discusses whether SkyNet like scenarios, where computers take over the world, are plausible. http://videos.syntience.com/ai-meetups/andiair.html

    • Thanks for the pointer to Monica Anderson, Bernd. I think that what she is talking about fits squarely with the bottom’s up view of intelligence. It makes total sense to me. I think this is also the beginning of some fundamental changes in the way our software works, moving away from explicit rules, to pattern recognition, where “understanding” is crucial. I like the way she breaks it out into the rational, the mundane and the mystical.

      You might be interested in “Radical Nature: The Soul of Matter” – it’s a bit of a slog to get through and is deeply philosophical. I’m reading it right now and it’s blowing my mind.

  7. An interesting follow up piece on gaming software that could read aspects of our emotions – within ten years.


    Thanks to Jennifer Sertl – http://twitter.com/#!/JenniferSertl – thanks Jennifer!

  8. “What would stop us from completely stripping them of their human soul?”

    Um, maybe Anonymous would if the technology got too powerful. But then the software/robots would defend against this and then maybe the hackers would create hackbots that would then go to battle against the Corporate  Softbots in scenario where software is battling software.  (Isn’t that already happening on a smaller scale with the antivirus programs running on your computer).

    Consider Asimov’s 1st law of Robotics:  A robot may not injure a human being or, through inaction, allow a human being to come to harm.

    Could being laid off  eventually count as injury? If so would the software optimize for the best possible outcomes for all humans or just shareholders and employees.  Code is law as they say, so we had better code not just our software but our society with all human outcomes in mind.