How to be tech in the tech lash – Ottoman androids and Pseudo-AI
AI and machine-learning are on the verge of becoming meaningless buzzwords and we could be deepening the tech-lash if we as startups are promoting technologies, we have not created yet, advocates COO and founder of legal tech Contractbook.
In the 18th and 19th century The Mechanical Turk a famous android toured Europe and America where it was presented as an automated chess-playing miracle. It replicated a traditional Ottoman sorcerer with robes, turban, and a pipe, and appeared to have an almost magical mechanism that enabled it to play a solid game of chess against any opponent. Obviously, such a thing attracted huge crowds and the hype was only enhanced when it out-smarted historical figures such as Napoleon Bonaparte and Benjamin Franklin.
The machine did not have any chess-playing abilities, though. Instead, it was a brilliant trick containing multiple optical illusions as an actual human chess player would be hidden inside the construction and controlling all the Turk’s moves. People were easy to convince back then since mechanical androids were the talk of the town.
Their complicated clockworks and human resemblances aligned perfectly with the popular scientific idea that the world was one big clockwork setup by God. Some saw the androids like scientific wonders that proved mankind’s supremacy, and others regarded them as entertaining wind-up birds. A significant group, however, feared the machines and regarded them as soulless, perverse, and bewitched threads to well-meaning and reverent citizens.
That is how it has always been with the newest technology. Fire, one of our first technologies, has the ability to both heat and burn which is symptomatic for technologies. Therefore, it has always divided people between the afraid and the enthusiasts. To be tech in the techlash, we have to remember that, and not distance ourselves further from sceptics.
Today, the modern tech-industry is packed with mechanical turks. The Guardian recently pointed out that we are witnessing the rise of “pseudo-AI”. Everywhere, tech companies are using humans to do bots’ work and even to pretend they are bots while doing it. Either while they are working on an expensive technology that will eventually automate the work, or because it just sounds cool when you throw AI and machine-learning into the mix. I get it. There is no doubt that it sounds cooler to have some complicated AI algorithm than to employ 50 Indian contractors.
The financial part of the question is crucial as well since AI and start-up budgets are really hard to combine. “It’s hard to build a service powered by artificial intelligence. So hard, in fact, that some start-ups have worked out it’s cheaper and easier to get humans to behave like robots than it is to get machines to behave like humans,” the author of the article correctly points out.
The practice is also referred to as “The Wizard of Oz technique” and was originally used in research and prototype development to simulate how a machine would behave in a given situation. It was often used when the technology wasn’t finalized but still needed testing of the interface.
These days, this totally valid method has become perverted into an actual business strategy. Chatbots operated by actual human beings, automated searches conducted manually by students and “photo recognition” made by contractors in Asia. While I am sure how widespread this is, I am sure that some Danish tech companies are pretending to have a technology they do not have. So sometimes the intelligence is more fake than artificial.
When AI makes you alien
A few weeks ago, the annual TechBBQ took place in Copenhagen.
At TechBBQ, it became clear that start-ups are becoming so fixed on AI and machine-learning that they are on the verge of becoming meaningless buzzwords. There are so many cool projects using AI out there, don’t get me wrong, but you don’t get to be a “data-driven AI-start-up” just because you use an algorithm. And frankly, we risk alienating people from the digital revolution if we keep dressing our products up in fake garments.
Artificial Intelligence is a really powerful metaphor with mixed connotations.
On one hand, it sounds high-tech, futuristic and incredibly smart. On the other, it gives one association to something clinical, and as something threatening to make you unemployed. I am not saying that it is not a good idea to associate your product with AI (it’s obviously working to create hype) but it could result in a major backlash in the long run if you are not being honest about your strategy. First of all, because future investors most likely will find out about your manual setup when they conduct a due diligence. Secondly, because people aren’t as ready for AI as one might think.
“The machines are all right!”
As COO in Contractbook which is facilitating a SaaS platform for contract management, I have experienced that convincing certain persons just to digitalize can be a challenge. Some parts of society still think that the cloud is something extremely vulnerable (without realizing that their Gmail is also cloud-based), and it also happens that people feel concerned about their digital signature becoming manipulated (without realizing that it is also possible to hack ink and paper).
We should take our time to explain and educate these people, and we have to be careful we are not deepening the tech-lash by alienating people with technologies that we haven’t even created yet.
AI, if done right, offers a huge potential: to automate manual hassles, make our lives easier, smarter and even longer as they can be used to do amazing things like detecting heart attacks.
In the end, we succeeded in creating a real mechanical Turk when IBM’s Deep Blue defeated chess world champion, Garry Kasparov. The machines are all right! So, let us be honest and straightforward – let us build a sustainable digital future. We don’t need the bullshit to do it!