React vs Machine Learning — the plight of non-tech companies
Despite all the hype and money around AI these days, what practical options exist for non-tech companies with established software teams? Setting aside the convincing argument that sometime in the near future all companies will be tech companies, in the here and now non-tech companies remain and so do the necessary tradeoffs — like having to choose between Facebook’s popular React library for building user interfaces and the resurgent AI subfield of Machine Learning.
Both paths require a significant investment of resources.
Facebook’s “Thinking in React” tutorial highlights the paradigm shift and the enormous initial effort required. It’s not uncommon to hear programmers say “my brain hurts” when learning to think in React.
For anyone who’s taken Stats in university, they’ll understand that the burden with Machine Learning is high. There are machine learning algorithms, Python data science, tensorflow and a litany of other tools, techniques and training programs to learn. Case in point: Amazon launched its AI service just today.
A Google Trends comparison of React (red) versus Machine Learning (blue) suggests more people are choosing to learn React over Machine Learning:
Is this trend warranted?
Don’t get me wrong, React is incredible. But in choosing it over Machine Learning are non-tech companies making the right choice? What are the opportunity costs of choosing React over Machine Learning?
It’s interesting to see what parts of the world are trending in the opposite direction.
India, Pakistan, Egypt and Iran are choosing Machine Learning over React? That’s curious.
Conversely, Russia shows virtually no trending for Machine Learning. Maybe it’s a translation issue? That’s curious, too. Especially considering Russian hacking of the US election.
Surprisingly, Silicon Valley and all of California solidly favour React, as well. You’d think the West Coast would have pockets of blue like Massachusetts and Waterloo given the caliber of universities and concentration of AI talent.
It’s not clear what these trends indicate: possibly that the majority of non-tech companies are making the same popular mistake; or possibly that Machine Learning is more difficult to learn and so less people are climbing this mountain.
Alternatively, if tech-centered universities — or countries on the periphery, where disruptive innovation tends to arise — are good predictors of future value then perhaps the trend towards React isn’t warranted.
What we do know is that major software companies are rapidly acquiring people and products related to AI. The motley crew of Facebook, Amazon, Google, IBM and Microsoft have joined forces — in spite of their intense competition with each other — to create the Partnership on AI. Even non-tech companies are hoarding AI talent.
Last month at the Dreamforce conference— with its whopping 170,000 delegates — Salesforce pushed their new AI product, dubbed Einstein. In addition to the spate of AI companies Salesforce has recently acquired, the cost of the naming rights alone suggests the importance of AI on their roadmap.
The futurist Kevin Kelly spoke at Dreamforce (23:03 minute mark here) about the first industrial revolution and how it resulted from taking ideas, adding artificial energy and 10,000 companies were born. Kelly continued: we’re now in the first inning of the second industrial revolution where ideas coupled with artificial intelligence will result in an even larger explosion of new companies.
For those of you old enough to have missed out on the dotcom boom, the AI revolution is an even bigger opportunity — and a second chance.
From this vantage, React seems incremental while Machine Learning is Google’s 10x thinking. The former improves; the latter disrupts.
To forego Machine Learning is likely the greater long-term risk. So do non-tech companies have the intestinal fortitude to embrace Machine Learning and buck the trend? Or is there some way to have our cake and eat it too?
