This article is part 2 of a 7-part series called “AI Zeitgeist,” where we’ll be mapping out the details of AI adoption over the next 10 years and explore the critical changes in the AI ecosystem that business leaders need to understand.
In this installment of the Zeitgeist series, we’ll be exploring the ways that the terms “AI” and “machine learning” will be used and discussed over time—from the buzzword of today (in the “Emergence” phase) to the invisible tech layer of a decade from now (in the “Dispersion” phase).
(Note: If you haven’t read the first article in the AI Zeitgeist series, you should familiarize yourself with it first. In that article, I explain what the 3 phases of “Emergence,” “Adoption,” and “Dispersion” actually mean).
We’ll start off by listing a few company names that probably sound funny to you:
- “International Business Machines”
- “Apple Computer Company”
You don’t know those names instinctively, but you know what the companies are called today:
What does “business machine” even mean? I imagine some kind of Rube Goldberg contraption with pulleys and gears and rolling marbles. It’s an antiquated term, to the point of literally sounding “funny.”
Today, “artificial intelligence” and “machine learning” are hot topics, as validated by the tremendous increases in search volume that these terms attract on Google since 2011, and by the fact that every startup under the sun stating that they’re using machine learning. Even if most of the time they’re lying, pretending to “do AI” still gets them more conversations with clients and with investors. Hence, the hype continues.
It won’t always be this way.
“Artificial intelligence” won’t have the same hype that it has today.
How many “internet companies” do you see in Silicon Valley, or Boston, or Tel Aviv?
The term “internet company” is laughably redundant. The internet just is, like sunlight or gravity. It is part of the ecosystem in which we operate, any company in existence must put up with gravity (even Bezos has to fight pretty hard to get those rockets into orbit), just like every global company in existence must put up with the internet.
The same will eventually be the case with “artificial intelligence” and “machine learning.” We’ll hardly talk about them at all. As “artificial intelligence” fades out of the lexicon of hype, it will become more accessible for business leaders, and the way companies will leverage AI for competitive differentiation will change.
Let’s explore how these terms will evolve in the three phases of AI in the enterprise: “Emergence,” “Adoption,” and “Dispersion.” We’ll start with today: the “Emergence” phase.
Phase 1: Emergence
We call it “artificial intelligence”, “machine learning”, “deep learning”—and in specific cases we speak of specific application types, such as “computer vision”, “natural language processing”, and the like.
We speak of it mostly with awe and excitement, often with naive notions of its capabilities and challenges.
These words add perceived luster and sophistication to the conversation about one’s company or product, even if most buyers and venture capitalists want to say that they’re too smart to be duped. Millions of businesspeople attend expensive events which can now add “artificial intelligence” to the event title, double their ticket price, and still attract large audiences.
Nobody—investors, businesspeople, tech vendors—wants to “miss out.”
Because “AI” is such a new and different term, companies in the “Emergence” phase are likely to build special, separate AI teams – and many of these teams won’t initially be embedded in a specific function (marketing, fraud, etc), but will operate between or on top of existing teams or departments.
Our technical advisor Marco Lagi (Tech Lead of ML at Hubspot, PhD in Physical Chemistry) describes the scenario well:
People have not yet realized that its a tool like anything else… it’s treated a bit like magic. People are trying to spin up departments for it… this makes sense as a first step… many experiments will be run….
But as the models get better and better you’ll see much more distributed teams… they will work alongside the computer scientists who develop the product… they would be PART of all the IT teams across all department.
As AI makes its way into the “Adoption” and “Emergence” phase, these new “magic” (Marco’s words) AI skills will work their way into becoming part of the fabric of various departments.
Because “AI” and “machine learning” are hot “can’t-miss-out” buzzwords, we see a lot of the following:
- Tech event companies can sell more tickets (or charge higher prices) by running “AI”-related events, often just by adding AI presentations to an existing, long-standing industry conference
- IT consulting firms are using the perception of AI competence to attract clients. Even if the projects they sell have nothing to do with AI, they compete in perceived smarts by referencing AI on their homepage and in their service offerings (read our article about the “Fake AI Rebrand” tactic that many such firms use).
- Professionals of all walks of life are taking charge of fledgling “AI”-related divisions and initiatives within their organizations, granting them the luster of “coolness” that comes along with having “AI” in their title and new presumed expertise.
This isn’t all bad. In fact, some of the newer AI events are pretty exciting and bring some powerful minds to present on real and meaningful AI-related business topics, and we win whenever these companies email us to help sell their event tickets (see: Advertising at Emerj).
Some of it is bad, however, and there are a number of lamentable aspects about the environment of hype that AI is embroiled in today. We can discuss the pros and cons of AI in the “Emergence” phase:
Upsides of AI buzz include:
- AI startups are swimming in venture funding, giving lots of new business models and ideas new life. This allows AI to evolve in the real world of business thanks to much-needed funding.
- Online courses, conferences, and universities are catering to AI-related education needs by building curricula and programs, granting more people access to data science skills at a beginner and expert level.
- Executives, leaders, and even mid-level employees (technical and non-technical) are getting somewhat familiar with basic concepts about AI and data science, which might foster the kind of “connective tissue” between subject-matter experts and machine learning engineers in the future.
Downsides of AI buzz include:
- Exuberant AI optimism and “awe” of the technology might create an extended period of let-downs as we’ve seen with previous tech trends (VR and 3-D printing come to mind).
- Organizational leaders are often distracted by AI concepts and exaggerated press releases, tying up attention in “use-cases” that often aren’t even real.
- Companies rampantly overuse the buzzwords of “AI” and “machine learning,” claiming product features and “AI-powered” capabilities, much of which is totally fake.
Fortunately, we’ll tip-toe out of the world of hype as AI becomes more broadly adopted.
Phase 2: Adoption
Only adoption laggards (small and mid-size companies, companies in developing markets, companies in obscure industries) will refer to today’s “artificial intelligence” ubiquitously as “artificial intelligence” in the “Adoption” phase.
Many aspects of AI functionality have become synonymous with what we now call “artificial intelligence.” Reasonably simple applications or capabilities of AI will be referred to as a functionality by itself, without direct reference to “artificial intelligence,” in the same way that spam filters (such as those used in Gmail) aren’t referred to as “artificial intelligence” today.
- Customer service software which automatically routes new support tickets to the proper department might not be referred to as “AI” as it is today (present AI vendors in this space include DigitalGenius).
- Email marketing applications which automatically send emails at a time which is statistically most likely to be opened by the recipient might not be referred to as “AI,” as it is today.
- Auto insurance products of all kinds will involve pulling data from social sources, IoT sources, and a variety of other sources, pooled into intelligent algorithms. This won’t be a special “AI application;” this will be par-for-the-course “auto insurance.”
These functions and features will just be. In large part, this disappearing reference to “AI” will be due to two main forces:
- The fact that many AI capabilities will not involve active human involvement (data cleaning and formatting, algorithm training, iteration, and tuning), and will operate as simple and intuitive tools like Google Maps, Amazon Alexa, or other everyday uses of AI.
- The fact that whatever we grow accustomed to no longer surprises us or seems worthy of attention. This has long been the case with the elusive definition of AI, but is clearly a truth of the human condition, painted in words by a great many sages:
“Things grow familiar to men’s minds by being often seen; so that they neither admire nor are they inquisitive about things they daily see.”—Cicero, De Natura Deorum
In an industry firmly in the “Adoption” phase, the words “artificial intelligence” and “machine learning” won’t carry an immediate, mystic power (read: Hype) as they do today. Rather, the terms in an “Adoption” industry will be pretty firmly understood.
Charles Martin, PhD, head of Calculation Consultants, explains a potential transition from the “Emergence” to the “Adoption” phase:
[AI] has to become completely comodotized so [businesses] don’t have to take any risk. Right now AI is still a science, and that means you need to take risks. You don’t know what you’re going to build; you don’t know if it’s going to work…
The average company needs to have something done in 3 months, and you can’t do a complex AI system in 3 months… The question is: ‘How will it be commoditized?’
It used to be that you would sell software to people, like Oracle [does]. But maybe everything will be in the Cloud, and if everything is in the Cloud, that means you need to pay Google or Amazon a fee everytime you want to use this stuff.
He goes on to say he doesn’t know how exactly how AI will become commoditized, but when it happens, the “Adoption” phase will be one where AI starts to become ubiquitous in certain sectors.
Phase 3: Dispersion
In the “Dispersion” phase, “artificial intelligence,” “machine learning,” or some new permutation of intelligent computer science may occasionally be referred to, but only in the development of new realms of machine capabilities that are wholly not possible with the intelligent software of the future.
By the “Dispersion” phase, all software will be “smart;” all software will recommend, take action, personalize itself. This will be downright expected.
Today we don’t think of a mobile website as something special; it’s just what a “website” is: it works on mobile. Software of the future will do what we need it to do with the least amount of effort on our part; it’ll just be what “software” is.
There will likely be a name for the “Dispersion” future we’re moving into. It might be “smart software,” it might be “software 2.0”.
I’m not sure what it will be called (and I don’t care to go about “claiming” the name), but what I do know is that any and all interactions with technology will simply expect them to “work,” and “work” will sometimes mean:
- Taking actions for users
- Recommending choices or actions for users
- Prompting users based on our preferences and goals
- Preemptively solving troubleshooting problems
As mentioned, we may continue to speak about “AI” when it comes to new and deeper functionality that present “smart software” can’t match, but all that becomes normal will no loner be seen as “AI.”
This has been week 2 of the “AI Zeitgeist” article series.
In the coming weeks, we’ll be exploring the following Zeitgeist topics, in order:
- The 3 Phases of AI in the Enterprise: Emergence, Adoption, and Dispersion (AI Zeitgeist 1)
- How “AI” Will be Discussed in the Future (AI Zeitgeist 2) <— You are here
- The Evolution of AI Talent and Training (AI Zeitgeist 3)
- The Increased Accessibility of AI in Business (AI Zeitgeist 4)
- Buying and Adoption Readiness for AI (AI Zeitgeist 5)
- The Changing Landscape of AI Priorities of Business Leaders (AI Zeitgeist 6)
- The Competitive Dynamics of AI – Now and in the Future (AI Zeitgeist 7)
Next week we’ll explore how AI talent, AI training, and hiring for AI roles will change over the next decade ahead (hint: It’s not just technical AI skills that matter in the future AI-enabled organizations, and hiring needs will change drastically).