Decision Laundering as a Service

4 min read Original article ↗

Today we are launching a new service for decision makers!

Do you

  • Struggle to justify your selfish decisions to the company board?
  • Feel uneasy about employee exploitation?
  • Find it too painful to decide on a negative diagnosis for your patients?
  • Want to use systematic bias to enable growth of your insurance business?

Now imagine you can have all those gains with none of the nagging from your consciousness or authorities!

Meet Decidr - an AI platform that allows you to scramble your decision chain by introducing a layer of Deep Learning between the decision-maker and the consequences of their decisions.

Here is how it works

  1. Input all your institutional data to our cloud. If your country regulation mandates an in-country storage, don’t worry - just run it locally through our DL training software and send the finished network it produces - that way it’s no longer ‘sensitive data’ but it encodes the same information - isn’t that neat?
  2. Input the decisions you want to be making - here we don’t just mean the questions - What if the system answered something that doesn’t fit your strategy? No, you also input the answers to drive that strategy forward!
  3. Now that you invested all that effort it’s time to pay us. How fast do you want the answers regurgitated back? How many nines of availability? We recommend spinning your intent on altruism, that should unlock that huge government subsidy that you will probably need to afford Decidr. Alternatively, we also offer packages for #startups - for you the service is free at first with the fees spinning out of control long after you exit. For a small additional cost you can access our #blockchain-based API which should allow you to project enough credibility to get through the next funding round easy-peasy.
  4. The system is online, the answers are coming back, people can get confused, but not you, because you already knew :-) Time to put the newly-found machine-enabled confidence to work, push the decision through other systems & society to unlock the benefits you always dreamed of.
  5. When the decisions ultimately come back through civil unrest, employee unions or plainly just weak performance of your company - they will hit an impenetrable wall of cold machine reason and incomprehensible complexity. We guarantee that nobody can scale that and attribute the blame. Corollary: If anything goes wrong and you come after us you will find out that this whole ad has been machine-generated and where human interaction was required, the machines made them do it.
  6. Rinse & repeat, potentially under a different legal entity.

Testimonials

Decidr is like a Tor Browser for social networks - the decisions come seemingly out of nowhere and our CFO loves the expense cuts it helped us to easily justify!

ML guru @BigCorp

We had a lot of finger-pointing going on in our bank and when the heads started to fall we knew something had to give. After we got on Decidr and shifted our assets to crypto everything got much easier, nobody can argue with digital meritocracy!

ML ninja @BigBank


Pretty scary right? You got through by rationalizing that Decidr is just a satire - and yes it is. However the bad news is that any company that works with Machine Learning has a little ‘Decidr’ at their disposal. And it doesn’t even require malicious intent, even though I’m sure we have enough of that in our industry as well. No, the plain old incompetence and corner-cutting is all that is needed to take your pre-existing biases & tendencies and bake it into your ML training in a way that makes the resulting model exhibit the same properties. Or even exaggerate them. If you want to read more about how AI/ML deployments backfire and what mitigation strategies one can use, here are some resources:

Let’s talk on Twitter!