By Anthony Scriffignano, Duns & Bradstreet.
I once had a great opportunity to have a “conversation” with one of the leading AI agents on the planet…a truly awesome experience. The particular information stack that had been curated into the environment included physical and natural sciences. I was given only two rules: I couldn’t lie (because the environment would learn from my question), and I shouldn’t judge the agent by how long it took to respond, since this particular instance was tuned to produce the best possible answer regardless of time spent cogitating. My first question needed to be a good one. It had to show respect (by not being too easy), be fair (no asking about something outside the corpus of knowledge) and hopefully it would be something that would demonstrate that I knew a thing or two about natural language processing and AI agents.
Carefully considering my first question, I came upon the following: “Why do birds fly South in North America and North in South America?” We all know that birds fly to where the weather is warmer as the season gets colder, so they fly towards the equator, but we know this in a very complex, nearly intuitive way. Part of what makes this question difficult for a computer is that the words “North” and “South” are used in two related predicates in different meanings. In the first predicate, south is a direction and north is a part of a proper noun, while in the second predicate south is part of a proper noun and north is a direction. After some time, the machine came back with a response: “migration.” I doubt that all of the nuance was considered, in fact I suspect that the second predicate was largely ignored by the AI agent, but nevertheless it produced a reasonable answer that most humans would consider correct... very impressive.
I expected the machine to fail, but this type of AI had improved subtly and consistently to a point where it could almost seem to be giving a human response. As I sit at my window in North America and watch the autumn trees turn to vibrant colors, I imagine the trees in the Southern Hemisphere budding with the light greens of Spring. I’m thinking about change… how everything is changing. And how that simple interaction with a very “smart” machine makes me think about how our interactions with technology, and our capabilities enhanced by technology are slowly changing what we do and how we do it. It seems the future is here in many ways. So what are some of the more subtle changes being brought about?
Auld Lang Syne: Jobs that don’t exist anymore
Not too long ago, coopers made barrels, cartwrights made buggies, and smiths did metalwork. Today, the Coopers, Cartwrights and Smiths of the world are more likely to be found in a cube farm than a farm with barrels and horses. Jobs like barrel-making and blacksmithing are largely relegated to nostalgia and a few very specialized craftsmen (and women). A closer examination shows that much more modern positions are slowly morphing or going the way of coopers, cartwrights and blacksmiths as well.
Technology jobs have undergone massive disruption from the onset. Phenomena such as outsourcing and commoditization have made once-lucrative jobs impractical or unavailable. Other jobs, such as key-punch operator, switchboard operator, and “computer” (a person who did manual calculations in the days before digital computers) have simply gone away as technology swept over industries such as telecommunications and accounting.
Disruption is ongoing in technology-related fields. Experts in tube technology and other electronics were disrupted by the advent of solid state technology, giving rise to new specialties such as integrated circuit design. Even “surviving” industries saw massive new requirements on existing technologies, such as power supply and battery manufacture. The difference in modern times seems to be that the pace of disruption is itself being disrupted. Skills that made us successful so far have become table stakes, as new roles and requirements emerge even before the competencies they displaced have stabilized.
What we do, and how we do it, are changing at a rate that belies simple compartmentalization into new “jobs” – the pace of change is requiring us to move from mastery of a few, well-understood competencies to mastery of change in a world of ongoing technology disruption.
The Constants in Change: Roles that evolve, rather than disappear
There are technology jobs that seems to evolve, rather than disappear. For example, my first full-time job was as a computer programmer. Today, programmers are called developers. This nomenclature isn’t simply a change in vocabulary. Back in the day, if we wanted to sort a list, we wrote a sorting algorithm based on one of several accepted methods. If we wanted to compress or encrypt data, we would similarly create our own code to do these component tasks. Today, of course, most applications are assembled with modules of pre-written code, which is either accessed or subsumed. Some code is still written, but the job of a developer is much more about assembling, integrating, and assuring forward and backward compatibility than in days gone by. There are standards, regulations, and conventions that must be respected. There are exigent considerations, such as cybersecurity and platform independence, which must also be considered. The job of a developer is not the job of a programmer. It is much, much more.
Another good example is information security. Since the early days of computers, there has been security. Initially, the only requirement was to control physical access to machines. As multi-user computing and time sharing became commonplace, account management, usually through simple passwords, came about to keep information private and compartmentalized. Of course, in today’s connected world, the job of IT security is constantly evolving, involving competencies in authentication, validation, site security, data retention and protection, and many other considerations. The physical threats have been augmented with threats that previously didn’t even have names, like malware, botnet attacks, and quantum hacking. Some of these threats have only begun to materialize, so the job of IT security will continue to move at a very fast pace.
Data acquisition and governance is another example of a technology-related field that is being disrupted, but which continues to evolve. Formerly a library-like function, the job of data curation and stewardship today includes rights management (permissible use), regulatory compliance, and many other considerations in addition to managing the content, quality, and character of information available to the enterprise.
Traditional IT-related roles may endure in changing times, but we should pay careful attention to how the requirements of such roles are ever-changing. There is very little room for status quo in these traditional roles.
Technology and Jobs of the Future: Will it even be called “technology”?
With phenomena, such as Artificial Intelligence, autonomous devices, and the Internet of Things, it seems technology will permeate most of our future day-to-day experiences. I’m among the people who don’t yet see the irresistible utility of an internet-connected refrigerator or a self-parking car, but I can’t deny the emergence of these, and many more tech-enabled devices. It’s likely that we won’t even refer to much of “technology” as such in the future, just as we don’t talk about “electronic” watches anymore, as the term might become largely redundant.
How will jobs in this near future change in ways we maybe don’t anticipate?
One example lies in my own field of data science. Today, data scientists are occupied with skills such as analytics, modeling, data visualization, problem formulation, and using data to answer questions and tell stories. I suspect that in the near future, systems will become so complex, and issues such as explainability (knowing how an AI agent reached a particular conclusion) and provenance (understanding by what right certain data was used to make a decision) will change the role of data science.
Data practitioners of the near future will be like clinical medical practitioners of today.
They will need the skills to examine complex systems, develop differential diagnoses (for problem resolution), and conduct trials to prove efficacy.
Just like human calculators gave way to digital calculators, digitization will also give rise to new, but subtly different jobs. Consider today’s economists, who take massive (and ever-growing) amounts of data into consideration to understand macroeconomic trends such as shifts in wealth and inter-regional monetary effects. During a recent interview, I had a very thought-provoking exchange with Mark Hughes, Program Editor for Al Jazeera, about how blockchain and cryptocurrency might impact economics. Mr. Hughes’ thought-provoking questions led me to speculate that in the future, we might have Cryptoeconomists, who specialize in the impact of digital currency in the borderless world of cyberspace. Does virtual money flow differently in ways that might challenge the basic tenets of economics? This potential new science is just one example of careers that might emerge from today’s disruptive technology.
We must be open to the emergence of new competencies and entirely new fields of study that are unknown to us today. Technology is not only disrupting, but also transforming the way we interact with information in ways that we are only beginning to understand.
Indeed, everything changes. In a world filled with technology that changes at a pace we are increasingly incapable of understanding, filled with amazing opportunity and ominous risk, it seems the one constant is our human resiliency. We must continue to reinvent, not only our technology, but ourselves. We must find the best in all that this change brings about in a time of abundant opportunity.
Bio: Anthony Scriffignano, Ph. D. is SVP, Chief Data Scientist at Dun & Bradstreet. He is an internationally recognized data scientist with experience spanning over 35 years, in multiple industries and enterprise domains.
Original. Reposted with permission.
- Why understanding of truth important in Data Science?
- How AI Learns What You’re Willing to Pay
- Machine Learning Engineer, Data Scientist – top emerging jobs in US