Even small businesses are using AI and machine learning
Data, technology and people are at their fingertips to make artificial intelligence and machine learning accessible to any business enterprise.
Certainly, artificial intelligence and its subdomain, machine learning, have gone through cycles of exaggerated expectations followed by disappointments. These cycles, which some call âAI winters,â go back decades.
For example, in the 1950s and 1960s, the United States government funded research into machine translation of languages. The hope was that the documents in Russian could be instantly translated into English. But in 1966, a report by the Advisory Committee on Automatic Language Processing, a government team of seven scientists, essentially killed machine translation research in the United States for about a decade.
“Computer researchers talk about a first computerized language translation system that has been tested by having it translate sentences from English to Russian and then back to English,” wrote Andrew Pollack, in a New York Times article, âTechnology; The computer as translator â, April 1983.
âFrom that went the statement ‘The spirit is willing, but the flesh is weak.’ The result came out: âThe vodka is good, but the meat is rotten. Likewise, the expression âout of sight, out of mindâ has become âinvisible, foolishâ.
Depending on the problem we’re trying to solve with artificial intelligence and machine learning, early researchers would probably have encountered problems with the collection and organization of data, the computational capacity to process that data (i.e. i.e. technology) and the availability of individuals to solve problems.
At various times and in different ways, this combination of challenges – data, technology, people – impacted the way AI-ML was used.
The situation is now different. Many companies, including most large corporations, are using AI-ML to solve problems as diverse as high-precision manufacturing calibration or managing a massive robotic processing warehouse (ask Amazon) .
AI-ML has solved many language related issues. For example, Google’s Word2vec algorithm can learn word associations and avoid misunderstandings such as “Vodka is good, but the meat is rotten”.
Google and other search engines use this concept when trying to understand the entities involved in a search query.
Data. Today, collecting, storing and sharing data is not much of a problem. Practitioners using legacy systems may find it difficult to transfer data from these older systems to a modern database. But overall, it’s not a big challenge.
Technology. The amount of computing power required to run AI or ML models was once prohibitive. Imagine trying to do machine learning translation in 1966 versus 2021.
I recently spoke with Jeff DeVerter, CTO at Rackspace Technology, the hosting and service provider. He said researchers in the early years of AI development had “the vision, but not the capabilities.”
Compare that to the current situation, when “anyone with a decent computer, graphics card, and GPU [graphics processing unit] who can offload some processing can not only make ML models work, but also make those models learn, âDeVerter said.
A single computer is not ideal for applying machine learning to, for example, a manufacturing operation or for rationalizing inventory. But the fact that machine learning is doable in the way DeVerter describes indicates that technology is no longer a barrier.
People. Historically, AI-ML required specialized data scientists who were in limited numbers. Even when the technology had the capacity for significant AI-ML, there weren’t enough experts to generalize. But around 2021, this is not the case.
âFirst of all, anyone who graduated from high school and had good math in high school has the option of spending six months or a year with classes. [from Jeremy Howard of Fast.ai] and become a practitioner of machine learning and data science, âsaid DeVerter.
âThere are a lot of really smart data scientists out there, and there are AI models out there that help run some of the science behind the scenes. You don’t need a data scientist on every project or certainly not a team of data scientists, âcontinued DeVerter.
People are no longer the obstacle to AI-ML.
AI-ML for all
Even small business businesses today use AI-ML. It is integrated with many or most software as a service platforms. AI-ML is also present in Google and the Facebook advertising platform, as examples.
The next step is for small and medium-sized business enterprises to use AI-ML to solve problems specific to their situation.
âIA-ML is 100% accessible to the smallest of businesses,â DeVerter said.