Artificial Intelligence Company Accused of Using Humans to Fake Its AI Tech

Artificial Intelligence Company Accused of Using Humans to Fake Its AI Tech

A shocking headline, right? Well... not really, 'pseudo-AI' actually happens a lot in the tech world.

The headline above was talking about the London based startup Receipt Bank - a digital bookkeeping platform touting AI and Machine Learning technology that can read through images of your receipts and digitise them - saving you hours at the end of the month usually spent manually submitting or type them up for your financial team or accountant.

Sounds great right? In fact, I use this very product and love it.

Anyway, the headline above accuses them of saying they use AI to turn the images to text but in fact, they actually outsourced the work to a virtual data extraction team to read the receipts and manually enter in the data.

I don't want to go into the ethics of this but instead the reason why this approach is used and if done correctly can yield great results.

This strategy is commonly known as a Concierge MVP in tech circles.

The point of the Concierge MVP is to simplify the product by replacing automated components with humans. Replacing tech and backend systems with humans front and center delivering the service.

Not only does the Concierge MVP save you a huge amount of time by skipping the product building, but it also puts you directly in touch with your early customers. This process allows you to understand their needs deeply, driving toward your fundamental question: “do people actually want what I’m offering?”

In Recipt Banks case - firstly, it's really hard to work on machine learning without a lot of data, and it's often hard to get that data without actual users using the product in roughly the way it's supposed to work. Secondly, it allows the company to prove there's commercially significant demand for the user experience that they hope to be able to eventually do with some algorithms/AI technology - saving them a considerable amount of time and money during the validation phase.

This strategy allowed Receipt Bank to test each of the three pillars of idea innovation:

  • Desirability: Do they want this? (problem validation)
  • Viability: Should we do this? (problem validation)
  • Feasibility: Can we do this? (solution validation)

I'm going to share some more concierge MVP examples shortly - for my post, I'm going to cover Zappos.

Product Strategy
User Testing