For many years, artificial intelligence (AI) and machine learning (ML) have been used as umbrella terms to describe highly technical processes that most people cannot understand. But more and more companies are transforming AI and ML from catchy buzzwords to real-world processes, integrating self-learning algorithms into everything from data analytics to mobile apps.
For many technologists, AI and ML are now essential skills in their repertoire. More and more technology jobs require AI and ML. do not have Knowing something about these technologies could prove harmful in a few years. But how do I get started with AI and ML? Are these skills? Really Is it worth learning?
We spoke to people already immersed in the field to understand how they got started, why they feel learning AI and ML is actually important, and where they feel the field is headed.
How it all started
Ian Patterson, CEO of cybersecurity company PluriRock, got into technology in a very roundabout way. “I started my career as a rock and roll roadie, designing fireworks and lighting, but stumbled into the world of technology when I moved to Victoria, British Columbia. I started working in the data science division of Terapeak, which was later acquired by eBay, and fell in love with the industry,” he tells Dice.
After that experience, “I decided to chart my own path in the field of machine learning and AI and started a data monetization company, Exapik, which was later acquired by FlashGate,” he continues.
James Kaplan, CEO and co-founder of MeetKai, was drawn to AI and machine learning through a different route. “For me, my initial fascination with AI was based on video games. I loved the idea of learning how the ‘intelligence’ behind the screen worked. It was often disappointing to see how simple the system was, but the fun of figuring it out was exhilarating.”
But Kaplan’s interests weren’t just limited to gaming AI. “While playing a lot of repetitive and arduous games from the early 2000s, I found myself thinking about how boring the games were. I ended up creating an AI bot to play the games. This made the act of playing the games a lot less fun, but it motivated me to get better at AI programming.”
Other companies are taking a much more traditional path into this emerging field. “I took a class on data mining and machine learning in college and was completely fascinated,” says António Alegria, Head of AI at OutSystems. “After completing my master’s degree in informatics and computer engineering, I started working on data-intensive technologies and products. My first job was at a large telecommunications company, looking at network packets passing through the network and automatically inferring the IT architecture of the entire organization.”
Chuck Everette, Director of Cybersecurity Advocacy at Deep Instinct, arrived a little late to the AI and ML party. “I realized the potential of a true AI framework when I was personally introduced to deep learning (a subset of machine learning) in 2018,” he says. “Having held cybersecurity leadership roles at multiple Fortune 500 companies and security vendors, I can honestly say I was amazed when I was shown the capabilities of deep learning properly applied to the cybersecurity space. I did my due diligence and dug into the technical aspects and nuances of this seemingly new technology.”
Everett added, “Over the years, I have used and been exposed to other security products that use machine learning, but deep learning has been, for lack of a better word, ‘amazing’ in its capabilities and accuracy.”
Are AI and ML skills in demand?
Some events have a sneaky way of punching above their weight. A headline-grabbing feature or product can make a language or skillset appear to be used much more than it actually is. But experienced engineers argue that this is not the case with AI and machine learning.
“AI is at the cutting edge of innovation and will continue to change the world as we know it,” Everett tells Dice. “To be honest, there is virtually no industry that is not already being impacted by AI in some way. Some of the most notable growth and expansion areas are healthcare, with diagnostic imaging and pharmaceutical R&D making heavy use of AI to streamline and quickly improve products and services.”
For example, AI can accelerate the research behind vaccines for things like COVID-19. The company, which helped develop Johnson & Johnson’s vaccine, collaborated with MIT researchers who leveraged AI and machine learning to guide the effort, according to Phys.org.
Jared Peterson, senior vice president of engineering at SAS, suggests this is more than just software. “Advances in deep learning, the computing required to enable those advances (such as GPUs), and the frameworks that make it all accessible have led to a renaissance in the world of computer vision and NLP. The pace of research and publication in these fields is staggering.”
Volodymyr Kuleshov, co-founder and chief engineer at Afresh, suggests that the convergence of hardware and software has big implications for the future. “The breakthroughs in AI over the past decade have been made possible by using specialized hardware to extend existing algorithms to large datasets, and this trend shows no signs of slowing down. We have seen breakthroughs in machine translation, language understanding, and other areas using scaled-up language generative models, and this is another area I expect to continue for several more years.”
Lorenzo Bavasso, director of data analytics and AI at BT Global, reminds us that while AI and ML are constantly breaking new ground, being leveraged for important purposes such as rapid vaccine creation, both are also powering the technologies people use every day. “The data industry is still evolving very rapidly in terms of the range of technologies, approaches and opportunities. Overall maturity is still far from complete. There is a huge gap between the best companies and the ‘rest’. Similarly, many AI/ML capabilities are still in prototype/pilot or “minimum viable” state at best, and are not being leveraged at scale. There’s a lot of focus on making AI/ML more “friendly” and less niche.
Babasso added, “For some evidence, just look at the landscape of the industry. AI/ML as a product is now part of most SaaS portfolios, the number of people working in the data and AI industry has increased significantly, and enterprise demand is exploding. Sophistication is still increasing, but I think there is an increasing emphasis on scale.”
The future of AI and ML
But where are AI and ML actually heading? Reactions have varied widely. Perhaps this suggests that AI and ML may have to dominate the technology for the next 5-10 years. If no one can agree on what the use case is, AI and ML can be a good fit for almost anything.
“If you asked me that five years ago, I would have said that our self-driving cars are talking to virtual assistant drivers,” Kaplan said. “My gut feeling is that over the next five years, we will see a movement towards bringing AI developed in the pure R&D world into consumer applications.” AI and ML cannot always be delivered in a consumer context, and improving speed and performance may take time. No one wants a search engine that takes 20 seconds to respond to a query.
Kaplan also points out that while AI will frame the way we work in the future, it could also impact how (and why) we hire talent.
I believe that within the next 10 years, augmentation/enhancement-based AI will be the default for most new hires. people will be hired according to their skills using AI increases efficiency. We are already seeing this in spaces where AI exists, such as drones and manufacturing. By 2030, it will pervade every white-collar and blue-collar job. It’s pretty exciting for me to think about. Think about how efficiently people work using smartphones. The adoption of AI in work will be orders of magnitude larger.
Alegria agreed with that assessment, saying, “Over the next decade, we will see startups applying AI to help organizations make day-to-day strategic decisions, such as algorithmic strategy and planning. This will not be about automated trading decisions, as we are already seeing in credit, lending, fraud detection, etc., but rather on strategic and impactful decision-making.”
Patterson also believes that “AI technology will be deployed at scale over the next five to 10 years. One of the reasons for that is that companies at the edge (like Apple) are using this technology in their hardware to enable speed improvements.” Mr. Patterson says: Even low-tech industries will begin implementing AI into their workflows and products.
Melody Yang, an iOS engineer at Apple, feels that ML modeling may soon become as common as JavaScript libraries. “Over the next five years, I think ready-made ML models for mobile apps will become easier to integrate and offer even more flexibility for customization. Current solutions include pre-trained models for transfer learning, and developers can use pre-trained models on their own datasets by fine-tuning some layers of the neural network or changing the activation function to match the desired output.”
“Progress is accelerating, but there are limits. Pre-trained models can only be used for tasks similar to the one they were trained for, and the dimensions of the new dataset must match the pre-trained dataset. More companies are focusing on making it easier for developers to integrate ML models into their apps. This means more solutions will be created for a variety of use cases and for developers of all experience levels.”
From simple solutions to healthcare services that overcome the pandemic, AI and ML are already deeply embedded in our technology-first society. Our experts fully agreed that both AI and ML will soon become so deeply integrated into our lives that tasks will become difficult to perform without some degree of assistance from trained models or “smart” assistants.
For technologists, this means that AI and ML are rapidly becoming core competencies, rather than little-understandable technicalities. These disciplines will determine how we accomplish most tasks in the near future. Ignoring that possibility is not recommended.
