Eratosthenes was the first to calculate both the Earth’s circumference and axial tilt. He did so by comparing the angle of sunlight at two different locations a known distance apart. On the summer solstice, he knew that at noon the sun would be directly overhead at one site located on the Tropic of Cancer. At noon on the same day, he had a collaborator note the angle of a gnomon’s shadow at another location. He was accurate with his estimation and it took place around 200 B.C.
In 2008, the global financial crisis was in full swing and I had teamed up with some college buddies to start a hedge fund. Our sense of timing wasn’t great because we started trading on September 2, right before the crisis was hitting its peak. Lehman’s Bankruptcy closely followed, and things only got worse from there. Back then, “deep learning” wasn’t something mentioned on LinkedIn profiles. If you wanted to use a neural network, you had to find an academic paper discussing it, decipher the equations for backpropagation, and then write your own code to get it working. The brainstorming session was intense, with plenty of coffee and whiteboards to keep us going. And the sandwiches from that local Italian deli didn’t hurt either. But I digress.
Nowadays, we have helpful tools like PyTorch, Keras, and TensorFlow that simplify the process of turning a neural network algorithm from theory into reality. Nevertheless, choosing the best marketing strategy or eliminating as much fraud as possible is not restricted by technicalities. The only thing limiting these activities is one’s creativity, originality, and ability to think innovatively – which allows for better questions to be asked.
The constant spikes in salary for professions such as developers, data scientists, and machine learning engineers have made many companies explore cheaper alternatives. However, these professionals demand high salaries for reasons that go beyond their ability to do a regression or automate chart generation.
They are invaluable because if Eratosthenes were alive today, he would be an exemplary scientist- one who has learned to code in order to better understand and solve global problems. Consequently, businesses would covet his skillset and employ him for these exact reasons. Today, no-code/low-code solutions are getting more and more advanced. However, we still need to make sure that they offer more than just superficial improvements. The challenge is ensuring that they are truly innovative and not just rehashes of things that have come before.
New “low-code” and “no-code” solutions pop up constantly, each promising to revolutionize business analytics. These tools claim they can provide C-level decision-makers with more insights than ever before. So, if these tools are the solution to all companies’ problems, why does the number of options seem so vast? When someone creates a better search engine, for example, it immediately becomes popular. The same goes for a stronger operating system or a catchier breakup song by Taylor Swift!
The reason why there is no one dominating solution in the no-code/low-code landscape is that we’re not using them correctly. Technology continues to evolve but we keep forgetting its intended purpose.
With each passing day, it becomes simpler to find answers to our questions. A calculator outperforms an abacus, a desktop computer is more powerful than a calculator, and GPUs scaled across the cloud allow for even greater computation. We can now access information and perform calculations more effectively than ever before. Even though it has become easier to find answers, posing good questions and designing strong experiments remain just as difficult. Even though AI, NLP, BCI, and ML are steadily improving, we still need human brains to run these technologies. In order to reach their potential, no-code/low-code platforms should be used as an aid rather than a replacement for the great questions that humans can ask. These tools do not replace brainpower; they enhance it.
When used in tandems with other tools, like wisdom and insight, low-code solutions can be useful. They often present excellent software that only gets better over time. This is something to get excited about. In the next few years, we will have low-code tools that help with experimental design, better inquiry direction, and even a modernized Eratosthenes measurement of the universe.
Given this situation, do you want to spend your time trying to avoid the brilliant scientist’s perspective? Or would you rather use low-code/no-code tools that can help find information and answers quickly, as well as involve humans who can ask questions insightfully? The combination of man and machine is what really makes progress.
