Since the founding of artificial intelligence as an academic field at the Dartmouth Conference in 1956, major strides have been made in machine learning, which has opened the door to possibilities that were once reserved for the plot lines of sci-fi TV shows. A major proposal emerged from that conference: “Every aspect of learning or any other feature of intelligence can be so precisely described that a machine can be made to simulate it”.
Fast-forward to 2018 – I am lucky enough to spend my days surrounded by experts innovating in the AI field. I’ve come to learn that AI has done more than just revolutionize our work, it’s changed the foundation of how we do it.
On the most basic level, we are now required to be adaptable, as this technology continuously challenges us to keep up with it. Beyond adaptability there are a few key principles I’ve noticed while working at Stradigi AI that would apply across nearly all businesses.
Let’s start at the beginning, before investing a tremendous amount of time into a project (AI or otherwise), you need to perform due diligence. Conduct research, perform experiments, back-test hypotheses and gather evidence as to whether your concept is worth investing in or not. This might seem obvious, but many enthusiastic entrepreneurs and executives will trust their conceptual implementation without being able back it up. Trusting your gut is a highly essential skill in the startup world, but it isn’t admissible as evidence when making your case. Ask yourself: “Is this proposal useful? Will it solve a problem? Can this improve people’s lives?” If the answer is yes, then you’re already halfway there. The next step is to collect enough data that proves that your concept can be translated into a viable solution.
If AI has taught us anything, it’s that it’s very difficult for people to argue with you if you have the data to back up your claims.
To quote the ever-quotable Yogi Berra “If you don’t know where you are going, you’ll end up someplace else.” Once you have the idea and the data, ask yourself – do you truly understand the problem you’re trying to fix? A deep and complete understanding allows you to identify the right path forward. Depending on your concept you must ask yourself key questions, such as: “Who is this for? Why do they need it? How can we make this essential?” Don’t fall into the trap of building your foundation on assumptions, because nobody wants to go into business with a solution made from a house of cards.
Machine learning is a great example of adaptability. It is remarkable due to its ability to learn and improve over time by taking examples and scenarios to build a path to a better and more effective solution. Follow AI’s lead and don’t trap yourself in a one-way approach; never stop evolving.
In the AI industry and beyond, “algorithm” has become a very common word. Yes, they’re fascinating and exciting, but if AI were a book, algorithms would appear in the second half of the story. Organizations shouldn’t assume they know enough to fast-forward through the beginning, before ensuring that their results truly reflect the solutions they are looking to achieve.
In the AI world, data scientists have the grueling job of cleaning data in order to ensure that the training models have the right information to learn from. This step is crucial because clean data is essential to reducing human and algorithmic bias. Besides data quality, data adequacy is a fundamental challenge to achieving accurate results. As good as clean data is, insufficient training will render confidence rates below an acceptable rate.
Don’t get distracted by “big shiny objects” – ensure both the quality and quantity of your data, otherwise your algorithm or your business concept will be useless in the long run.
When we hire experts to develop cutting-edge technology, we want them to focus on what’s most important: building solutions that are able to process information (or data), learn from it and then produce an output that is relevant to a client’s needs.
There is a common fear: artificial intelligence is improving our lives, but it may eventually replace people’s jobs. While the chances of this are quite high, much like in the case of any other economic revolution, some jobs will become obsolete, while others will be created. AI is a tool created by humans for humans, and it can be used to maximize our collective productivity. When you think about the work environment you are providing for your employees, AI has the ability to augment productivity beyond giving them benefits, flexible working hours and a competitive salary. It has the potential to remove the more mundane, repetitive and energy draining responsibilities of our work days, allowing us to be more creative and productive. Perhaps one day, this increased productivity can result in the 4 day work week.
To ensure that we are applying this principle — of maximizing human productivity, without trying to replace anyone — we need continuous input from users and diverse stakeholders to ensure that there’s equitable representation. AI is teaching us a great deal about our humanity, and that we can use technology and our work environments to augment the human experience.
According to Phil Jackson, “The strength of the team is each individual member. The strength of each member is the team.” Right now, the research scientists are, of course, key players in developing the AI technology. Looking around our office, you would notice software developers who are responsible for building applications, the design team that creates the user experience and the sales and marketing teams that put the product out into the market, relaying client and user feedback back to the appropriate teams. Every moving part is necessary to bring a project to term and the benefit of diversity in areas of expertise and backgrounds gives us the ability to see things through multiple lenses.
In order to be inclusive, you must give people the opportunity to learn and adapt their skills, but you also have to value the human element which plays one of the biggest roles in your success story.
Make sure you not only have data to backup your instincts, but that it is good, clean and unbiased. Before you go out trying to solve your problem, ensure that it’s well-defined and don’t skip steps. Your success will depend on a well thought-out strategic plan with a strong tactical implementation. Above all, put people first, whether this means building an environment that allows your teams to thrive, or including diverse stakeholders in the process to ensure human-centric development.
Interested in starting your AI journey? Contact us today.
