By Justin Bercich, PhD, Head of AI, Lucinity
Our entire lives, both inside and outside work, are dictated by the decisions that we make. In the main, we’re hardwired to subconsciously learn from our mistakes, to avoid bad decisions and to question how we’d improve our decision-making if faced with similar scenarios in the future.
And that’s exactly the same concept for machine learning (ML). AI (artificial intelligence) brains are, by and large, programmed the same way as a human brain. Advanced AI and deep learning are built to learn from human decisions, ask the same questions and reinforce the same principles. And the more seamlessly human that AI becomes, the more we can connect and relate to this incredible technology and the more we can trust it to sharpen and improve our decision-making and, ultimately, our lives.
Open source will level the playing field.
Put simply, a machine-learning algorithm watches—identifying the good decisions and learning from the bad ones—just like a human brain. And that’s the exhilarating maturity curve we now find ourselves accelerating along, one built on ML algorithms and AI models that are beginning to operate, inexorably, in the same way that humans do, through positive and negative re-enforcement feedback.
It’s no secret that technology, as a whole, has become more available, accessible and democratised. One of the main reasons AI and ML have been able to continue their relentless march is because of specific open-source mathematical software such as Tensorflow (deep learning) and Kubernetes (distributed computing), which have made data science infinitely more efficient and effective. The more people who become fluent in Tensorflow and Kubernetes, the more ideas and innovations will flow and flourish, and the more advanced artificial intelligence and machine learning will become.
It means that machine-learning pipelines can now be industrialised, operationalised and commercialised. Traditionally, machine learning was a process that took place offline, with models updated using data outside production. Now, the machine-learning pipeline is built on algorithms and models that learn efficiently as data flows through the system. Brands that have cracked this “deep learning” code will understandably keep their cards close to their chests, because it’s so valuable.
In a nutshell, machine learning is beginning to increasingly resemble a conveyor belt. You receive data, you make transformations, you make a prediction, and then you learn from it. Your machine brain is always learning from new insights given to it, just like a human brain.
Human AI: Machine brain, meet human.
Another reason the use of AI and ML has increased exponentially is the way in which their insights are presented and the human trust and connection that garners. Just five years ago, only data scientists were able to decipher and extract meaning from machine-produced data. After all, a machine wrote it. It was intangible and had no connection to a human’s thought patterns.
But now, the way technology has evolved and the way we present these algorithms and models have become infinitely more personal. We can tangibly connect with and relate to machine-learning outputs. As soon as you start seeing parallels between the way a machine acts and your own personal behaviour, you begin to treat the technology as a peer you trust, and you’ll derive so much more value from it than before.
For example, in order to help banks and financial institutions combat money laundering, we send the human feedback we gather from our investigators straight back into the machine-learning system. This means the insights go beyond merely flagging whether an incident is money laundering or not. This deeper level of human-led learning includes the rationale behind exactly how that investigator came to his decision, where he spent his time investigating and the analysis behind his narrative.
This increased sophistication represents the gradual shift we’re experiencing as AI and ML evolve from an intangible ideology into practical execution. The key here is that AI is beginning to solve single issues across an entire organisation, overcoming tangible problems—from the front lines up to management—to which everyone can relate, not just a data scientist interrogating complicated reams of data. Right now, 63 percent of people don’t even realise they’re already using AI technologies.1 But, as we increasingly see them solving relatable problems—think of your Spotify song recommendations or Google Maps re-routes—our understanding and trust of this technology will strengthen.
And it’s a self-fulfilling prophecy: The more we trust this technology, the more we interact with it on a human level. and the more human it becomes (in a non-scary way, of course). And, in turn, AI and ML become less reliant on forced behaviour and pre-determined algorithms, instead absorbing natural human behaviour, which will invariably increase the quality of its output and ultimately bolster that trust even further.
References
1HubSpot: Artificial Intelligence Is Here – People Just Don’t Realize It
