Digital Analytics + Marketing Career Advice: Your Now, Next, Long Plan

The rapid pace of innovation and the constantly exploding collection of possibilities is a major contributor to the fun we all have in digital jobs. There is never a boring moment, there is never time when you can’t do something faster or smarter.

The tiny downside of this is that our parents likely never had to invest as much in constant education, experimentation and self-driven investment in core skills. They never had to worry that they have to be in a persistent forward motion… sometimes just to stay current.

This reality powers my impostor syndrome, and (yet?) it is the reason that I love working in every dimension of digital. We are at an inflection point in humanity’s evolution where in small and big ways, we can actually change the world.

With that context, this post is all about career management in the digital space. Like this blog, it will be particularly relevant for those who are in digital analytics and digital marketing. I would offer that the higher-order-bits in each of the three sections will provide valuable food-for-thought for anyone in a digital role.

The post has three clusters of advice. The first two are from editions of my newsletter, The Marketing – Analytics Intersect (it goes out weekly, and is now my primary publishing channel, sign up!). The third section was sparked by a question a friend who works at a digital agency asked: Will I lose my job to automation soon? (The answer was, yes.)

The Now section provides advice on how investing in growing your Analytical Thinking will contribute to greater success in the role you are in. The Next section provides advice on what you should be doing to invest in yourself to get ready for the depth and breadth change Artificial Intelligence is going to bestow upon us (regardless of your business role). The Long section shares a thought experiment I want you to undertake to figure out your career three years from now.

One more change reflective of the times we live in… Your employer used to be responsible for your career, this is for the most part no longer true. Your employer would send you to trainings to help push your career forward, this is for the most part no longer true. Your employer/manager would help you figure out the skills you can develop, this is for the most part no longer true. It is now all on you. Hence… Take control.

Ready?

The Now Career Plan: Analytics Experience vs. Analytical Thinking

Check the requirements listed in any digital analytics job and you'll notice a long laundry list looking for analytics experience.

Years of having used tool x. Years and years of practice with R or "Big Data." Years of proficiency in analyzing m campaigns for n channels resulting in production of z reports.

When you go to the interview, the hiring company will proceed to ask questions that test your competency in the listed job requirements.

This is normal.

Reflecting on my experience, it is not sufficient.

Test for analytics experience AND explore the level of analytical thinking the job candidate possesses.

Analytical thinking is 6,451 times more crucial in the long-term success of the candidate and the value they'll add to your company.

Analytical Thinking: Skills, Interviewing, Value.

Analytical thinking is a collection of skills.

It is creative problem solving. It is working systematically and logically when dealing with complex tasks. It is exploring alternatives from multiple angles to find a solution. It is a brilliant evaluation of pros and cons, and achieving the balance that is right for that specific moment. It is always knowing that the answer to what's two plus two is always in what context? It is being able to recognize patterns. It is knowing that every worthy life decision is a multivariate regression equation (hence the quest to identify all the variables in that equation and their weights). It is the possession of critical thinking abilities. And, most of all it is being able to seek and see the higher order bits.

Beautiful, right?

If I have the immense privilege of interviewing you, expect us to spend a lot of time on the elements mentioned above.

One sample strategy: Expect that I'll ask open-ended questions (If a company has 90% Reach on TV, why the heck do they need digital?). Then, regardless of what you say I'll politely but forcefully push back, to explore the depth and breadth of analytical thinking you bring to the table.

If you hire strong analytical thinkers, of any background, you are hiring people who will be adaptable, who'll grow and flex with your organization and needs. They'll have the mental agility to think smart and move fast. They'll ask child-like simple questions that'll lay bare your complex strategic challenges. Hire them. And, if they don't know tool x… You can teach them which buttons to press.

Caring and Feeding Your Analytical Thinking.

If you are an analytical thinker, there are many ways in which you can keep feeding and stretching the synapses in your brain. There is always more you can learn.

In a business context, request an hour to talk to people three levels above you in the organization. Ask them what they worry about, ask them what they are solving for, ask them how they measure success, ask them what are two things on the horizon that they are excited about. So on and so forth. You'll see things very differently, and you'll think very differently when you go back to work.

I'd mentioned being able to look at every situation from multiple angles. (Think of the famous bullet time scene in the Matrix.) Hence, a personal strategy of mine is to look well outside my area of expertise to help me improve my analytical thinking capabilities.

I'm love reading decisions of the US Supreme Court. SCOTUSblog FTW!

The Supreme Court deals with situations that are insanely complex – even when they appear to be stunningly simple on the surface. There are so many lessons to be learned.

My favorites are the ones I massively disagree with. Citizens United is one such example. I could not possibly disagree with it more. Yet reading through the deep details helped me see the multiple facets being explored, the reasoning used by the other side. I learned a lot.

I go in open-minded, and at times have my mind changed. A good example of this Justice Scalia's opinion in Gonzales v. Raich and the use of the Commerce Clause. And, he was not a man with whom I have overlapping views on anything. I appreciate him stretching my mind in this case.

Optimal Starting SCOTUS Starting Points.

If you would like to pursue my personal strategy, here are a collection of cases to use as starting points.

Some cases are very dear to me, I truly love them, there is a lot to learn from them as you explore the back and forth of the debate, the majority opinion and the dissenting one (or ones).

Loving v. Virginia is close to my heart, it is the reason I can legally marry my wife. It was just 50 years go!

Obergefell v. Hodges brought immense to our family as we celebrated the right of all Americans to marry. Justice Kennedy's opinion is a thing of beauty. And, it is also useful to read Justices Scalia and Thomas' strong and powerful dissents.

Texas v. Johnson said that prohibition on desecration of the American flag was a violation of the right to free speech. Of the many wonderful things about America, the First Amendment is at the top and distinctly unique. The court looked beyond the jingoistic distractions the flag always attracts, and protected what's critical.

As I'd mentioned above, there is much to learn from cases that are heartbreaking

Dred Scott v. Sandford held that African Americans, free or slaves, could not be considered American citizens and undid the Missouri Compromise. It contained the infamous quote "[black men] had no rights which the white man was bound to respect."

Buck v. Bell is perhaps the one that is a deep, deep source of pain for me, it a decision that still stands. The court upheld forced sterilizations for those with "intellectual disabilities" and contained the despicable phrase "three generations of imbeciles are enough."

Korematsu v. United States, legalized the shameful internment of American citizens with any Japanese ancestry. It is still on the books, and places extraordinary power in the President of the US to do what they want to people who might not look like "Americans." People like me.

Each case, regardless of if I agree with the opinion or disagree, helps push my thinking. It makes me a better analyst, a better employee, a better start-up founder.

I've added a differentiated collection of links above to take you to sources, I hope they'll help feed your analytical thinking.

For the Busy Human On The Go, An Alternative.

Given everything above, I absolutely LOVE the More Perfect podcast.

Jad Abumrad and his team are magnificent storytellers. For each episode, they take one case and explore it from multiple directions. They are entertaining, engaging and deeply informative.

Season one covered seven scintillating cases. I found the episodes that shared how SCOTUS was formed and got its power amazing.

Season two kicked of with… Korematsu! I thought I knew all angles of this case. Yet, towards the end you'll hear two loud silences in a conversation with Judge Richard Posner. Make sure you hear what he says. I have profound respect for Judge Posner, he is brilliant. And, in those two moments, he both made me deeply uncomfortable and appreciate complexity.

More Perfect on iTunes, Stitcher, Google Play.

Bringing It All Back To Analytics.

The latest episode (as of Oct 11th) is "Who's Gerry and Why Is He So Bad At Drawing Maps."

The problem is simple. In Wisconsin Republicans in power massively gerrymandered voting districts (something the Democrats also do when in power). Unlike the past where little sophistication was applied, this time sophisticated algorithms and computers were brought into play. Resulting in more effective gerrymandering.

End result: Democrats won 53% of the votes but only 39% of the seats.

You might think: OMG! CRAZY BEANS! What happened to one person one vote!

Well, the case was heard by the Supreme Court last week. And, everything's quite complicated (analytical thinking!). Listen to the episode for that.

What's even more material for us is that Justice Kennedy wants to know how can he figure out that a district has been "too" gerrymandered. There is no real standard, nothing the Justices can use.

Math to the rescue!

Nicholas Stephanopoulos and Eric McGhee created an Efficiency Gap formula to assess how bad the gerrymandering was. (More here, PDF.)

I won't spoil it for you, let Professor Moon Duchin explain it to in the podcast. It is a thing of beauty.

You'll learn how to create smarter formulas in your job, how to solve complicated and ambiguous challenges with simple assumptions, and how to not to grow too close to your formulas – rather evolve them over time to be smarter.

In 23 mins, it will make you a better Analyst.

If you follow the overall guidance in this section, you’ll continue to invest and grow the one skill you’ll need in every digital career: Sophisticated analytical thinking.

The Next Career Plan: Prepping For An AI-First World

Even with all the hype related to all things Artificial Intelligence, I feel people are not taking the topic seriously enough. That the big, broad implications for the very near future are not causing us to sit up, take notice, and change our strategies (personal and professional).

Or, maybe I'm just too deep into this stuff. :)

I had two big ah-ha moments that have changed my view if humans can be competitive in any field compared to what technology will spring forth. I call the two elementsl Collective Continuous Learning and Complete Day One Knowledge, they are harbingers of exciting possibilities for what we can do with AI (and it to us).

For more detail on that, and if humans are doomed (yes, no, yes totally) please read: The Artificial Intelligence Opportunity: A Camel to Cars Moment

The topic of AI is vast, and I’m not even including all the layers and flavors. The more I learn, the more I realize how little I know. My heartfelt recommendation is that every professional should be curious about AI and try to stay abreast with as many new dimensions as they can. After the first few months, you’ll find your own sweetspot that’ll catch your fancy.

Here are the collection of books, videos, people and learning opportunities from my sweetspot…

Books.

I want to recommend three books. None focusses on digital marketing or analytics. Each tackles humans and the possibilities for humans. Hence they’ve had a profound impact on my thinking about humanity’s future (and via that route, my career plans).

1. Homo Deus: A Brief History of Tomorrow by Yuval Noah Harari.

The span of Mr. Harari's thinking is truly grand, and he's a great storyteller. I am less pessimistic than Mr. Harari about the 300 year outcome (as you'll read in my post above on AI), but he's influenced my thinking deeply.

2. Superintelligence: Paths, Dangers, Strategies by Nick Bostrom.

AI will birth numerous incredible solutions for humanity, but the most magical bits will come from Artificial General Intelligence. Some people think of it as Superintelligence. Mr. Bostrom does a fantastic job of exploring the possibilities. Let me know if you get scared or excited by the end. :)

3. Life 3.0: Being Human in the Age of Artificial Intelligence by Max Tegmark

I love the way Mr. Tegmark writes, and there is something magical about his ability to distill all living things, you, me, watermelons, to up quarks, down quarks and nand gates! I was so inspired by his writing that I wrote to him my personal prediction for humanity looking 300 years out.

Videos.

Current development of Intelligence is in silos, I'm glad when someone pulls all the experts from around the world in an attempt to guide humanity's efforts.

The Future of Life Institute hosted a conference in Asilomar in Jan 2017 with just such a purpose. The entire list of videos is well worth watching, prioritize the individual ones: Beneficial AI 2017

If you can only watch one…

1. Science or Fiction?

The content is great and it is pretty amazing to see these crazy brilliant group on one stage.

There is one other video I want you to watch, from the 2015 edition.

2. Robotics, AI, and the Macro-Economy

There is mostly a negative vibe about the combination of robotics and AI. The brilliant Jeffrey Sachs systematically presents context you'll be glad you've heard.

There is a ton of video content on YouTube. A go to source for me is whoever is curating the Artificial Intelligence AI channel.

People.

In any space that is having the kind of exponential growth like AI, your best bet is to find people who trust and listen to what they are saying/doing.

We are blessed with a ton of experts, practitioners and futurists. I encourage you to curate your own list.

Here are the ones I follow as closely as I can: Sebastian Thrun, Jürgen Schmidhuber, Demis Hassabis, and Andrew Ng.

I watch videos of all their talks on YouTube or tune in to livestreams of their presentations. I read articles they write. I have alerts for them. Luckily they are so darn busy, they pace their public speaking/writing. :)

You can follow their work using strategies you currently use for others you stay in touch with.

Learning.

If you are slightly technically oriented and would like to start your journey of acquiring technical knowledge in the space, Udacity is a great place to go.

All three of these courses are free:

1. Intro to Machine Learning

2. Machine Learning

3. Deep Learning

If you are deeply technically oriented, you already know where to go and don’t need my pointers!

I’m sure you’ll notice I’ve not given you specific advice for your next career move. One reason: We are in a moment where each of us has to know all the changes coming, all the possibilities arising, and then figure out that answer for ourselves.

The above books, videos, people and lessons will help you discover the right answer for yourself.

The Long Career Plan: Automation & Your Value To A Company

People are scared of automation.

It is logical. The AI revolution will bring a ton of automation that will eliminate current white-collar jobs in large numbers.

Yet, by the end of this thought experiment, you might see that looking out over the nest 25-30 years, we can deal with automation (/elimination of our current jobs).

This thought experiment is for both Marketers and Analysts.

Get in front of a whiteboard. Draw a decent size square box on it.

Today, almost all the work you do is inside that box.

For Digital Marketers, it is finding keywords or websites, setting targeting parameters, building ads, setting bids, adding rules, building landing pages etc. etc.

For Digital Analysts, it is creating data collection mechanisms, writing queries, creating reports, doing segmentation, creating rules, identifying business focus areas based on data etc. etc.

Here's the thought: If tomorrow everything you currently do, inside that box, is completely automated… What's your value?

Pause.

Think about it carefully in terms of personal implications.

For the bravest among you, think of what's the value of your Agency/Company.

If you are anything like me, you are super-scared. Some of you are likely super-excited as well.

Don't be scared. Take action.

It is not as crazy as you think to envision that you could be completely automated out. In small pieces this has already happened.

Media example: Campaigns to create, target and deliver results for driving app downloads is now almost entirely automated.

Analytics example: There are already buttons in your tools that automate finding of anomalies in your data that your leaders most need to pay attention to. Eliminating the need for the known knowns and automatically providing the known unknowns and unknown unknowns.

An example that combines the both for even more effective automation: With smart creative, smart bidding, and smart targeting there is no need for any human to touch AdWords or soon a whole lot of your Display campaigns. The results of Data Driven Attribution modeling, which use data from *all* digital campaigns, can now be directly plugged into AdWords which means without any reporting/analysis the platform will automatically optimize for the highest profit for your business – with no human involvement. This is not the future, it is Nov 2017.

Back to the whiteboard.

On top of the box with the stuff you do, write the word Automated.

Ponder now what's your value.

You'll see there are two areas where you can add value. The area before the box, the area after the box.

If you are a Marketer…

You can shift to taking more ownership of the inputs that go into your current job (which remember is now automated). Shift to a responsibility that requires a deeper understanding of your Prospects and Customers at a human level. Now, because of that beautiful knowledge, take ownership of the entire process of identifying the optimal creative assets required for any great Marketing campaign. Then, step up and move to the other side of the box… Own the use and deployment of large scale machine learning services to understand every human, which results in creating the simplest most meaningful experience across all digital touch-points. And then… I'm taking you so far away from your current box… expand the outcomes you own from just the transactional to building deeper years-long beyond-pimpy relationships with your customers.

And suddenly…

You hate the freaking box you are in as a Marketer today. You want to expand your responsibility to own these deeply meaningful things that Machine Learning and our Deep Neural Networks won't touch for a while. You want to feel the true joy that comes from doing meaningful things like figuring out how to build relationships or unleash the full and beautiful power of amazing creative (in ads, in apps, on sites, in products), and so many more exciting things that you were born to do.

Now, you are not scared about automated. You can't wait for your current job to be automated away.

:)

I have the above scenario and the wonderful possibilities for Analysts as well. It is also very exciting, as you’ll discover when you do the whiteboarding exercise for yourself.

Now. I totally get that your entire job is not getting automated tomorrow. But, I suspect you'll be surprised though how fast that is coming. For Nurses. For Truck drivers. For Baristas. For… Everyone. Collect a handful of the smartest people you know, draw a box on a whiteboard, have a discussion.

This thought experiment is just one way to think through the implications of what’s ahead of us. In my blog post on the artificial intelligence opportunity, you’ll see another way I framed how to think this through…

The above framing is a bit more in the higher-order-bit spirit.

I recommend the thought experiment. When you’re done: Step one, have a plan. Step two, execute. Step three, joy. Step four, follow the advice in section one (Now) and section two (Next) of this blog post and start investing in the personal growth you’ll need to move to these new more joy-inducing meaningful jobs.

Your career is in your hands, and I deeply believe it is going to be bright. Seize the moment!

As always, it is your turn now.

Considering the Now moment, is there something unique you do to invest in growing your analytical thinking capabilities? How are you preparing for the Next moment, who are you reading, who are you listening to? Considering the next 25 years in our space, how far do you think automation will go? How are you approaching your personal evolution with the Long moment horizon in mind? How about your company’s?

Please share your unique perspective, challenges, and solutions via comments below.

Thanks.

P.S. I've touched on the topic of career paths and career management in earlier posts. Here are a couple you'll find to be of value:

Comments

  1. 1
    Tom Lawson says:

    A very provocative collection of ideas Avinash.

    My first realization is that the world is in front of me, yet I'm not seeing anything. You look at that same world and see patterns that are telling you something completely different. I appreciate having my eyes opened.

    I am curious what you see before and after the box for Analysts. I promise to do the exercise with my team, but if there is anything I've learned from this post it is that you are going to come up with something completely different.

    Thanks.

    • 2

      Tom: Because I know Analysts so well, being one myself, it is likely a longer post.

      Briefly though… Analysts focus too narrowly on one part of the food chain. On the right side of the box for example are a whole host of things related to creating a sense of urgency for action. One simple strategy for that is leveraging smart predict analytics around what data is saying you should do. That's just the data side, there are a whole host human/emotional strategies to create a sense of urgency.

      Most analysts don't do this at all. For now, ML will likely need humans to do this. Gainful and joyous employment for you and me!

      Avinash.

  2. 3

    A great post Avinash and love your vision and simple framework of the whiteboard, very clear to understand!

    I have already tried to sketch out the possible skills that fit in "before" and " after" sections of the box you mention and within these before and after sections, and my take on this is that there will continue to be opportunities that exist for the more softer skills as Seth Godin refers to: https://itsyourturnblog.com/lets-stop-calling-them-soft-skills-9cc27ec09ecb

    But I am curious to understand what and how these soft skills (if that is indeed the case that they are of value?) can be turned into opportunities to pivot your career and the types of roles/jobs they will create?

    Can you picture what these jobs would be and how would you define an ethics centric or empathy centric job?

    All the best
    Simon

    • 4
      Trotte Boman says:

      An empathy centric job is things like working at a childrens daycare.

      For analysts I think you will have to lean more towards "creativity" and "separate good data from bad data". For every step ML and AI takes, there will be 10 new models applicable to the problem domain you are in. Being able to understand which of the available to apply with the available data and how to present it to the machine will require great domain knowledge for a long time to come.

      The even better analysts will be able to look at the best model and understand how to tweak it to be even better for the given problem. I believe that the further we get with ML the more we will be able to rely on creativity rather than excellent skills in mathematics for improving the existing models

      • 5

        Trotte: You are completely right about the reliance on us for creativity of various kinds.

        Over the long term it will be interesting to see what facets of creativity stay with humans. In the lobby of the PartnerPlex in B40 at Google in Mountain View, there is a piano that, automatically, plays music written by an AI. I am always stunned at how good the music is. Not yet Beethoven, but better than so many others that are good at this stuff!

        Avinash.

    • 6

      Simon: A whole bunch of soft skills will be valuable in the interim (during the phase ML continues to become ever more effective at doing more in more jobs), some might even be relevant even on the long run.

      Here's what I want you to consider, say three decades out…

      Today in Marketing there are fifteen or so departments, for larger companies it might also involve two or three agencies. Across all these layers, you need decisions made and action to be taken on those. Lots and lots of soft skills required (often sadly even more than data).

      In a couple of decades lots of the above points will be connected via automation and intelligence. You simply won't have humans involved between information assessment – identification of influencing actions – begging the right leaders/departments – taking action. No humans, no softskills.

      Some roles will still remain where humans will still need to be involved, requiring soft skills.

      Avinash.

  3. 7
    Vinay Sharma says:

    I am both scared and super excited at the same time.

    Beautifully put together different aspects of future of digital marketing and analytics…

    World is changing because of AI and Machine learning and it is changing fast. Most important skill we need is to unlearn fast to learn new…

    Thanks for sharing your wisdom.

  4. 8

    I think mostly importantly we need to evolve ourselves daily. Knowing the things happening around us is essential.

    I was in a box doing repetitive tasks but recently I started learning new topics through MOOCs like edx coursera udemy etc. They have excellent topics for AI, deep learning, UI UX strategy, analytics strategy & more. Also I started exploring new startups & their ideas.

    The world seemed much bigger than the tasks I was doing. Because of learning new topics & putting out an hour or two daily for learning new things, I could evolve myself to the tasks where AI or ML might not help.

    I guess in coming years human human interaction will be the most important skills to have.

    But still I am not skeptical about AI & ready to evolve according to it.

  5. 9
    Jim Moore says:

    It brings such joy to find a fellow lover of the Supreme Court. I am a voracious consumer of all things scotus. Now I can show your blog post to my boss and do all my reading during work hours. :)

    I have not been consciously leaning into developing my analytical thinking nor the broader implications on AI. I would argue that your recommendations in Next are a way of pushing our analytical thinking forward.

    I appreciate this wake-up call Avinash. You rock.

  6. 10
    Dushyant Joshi says:

    Thanks a million for this Avinash.

    Best regards from New Zealand. Dj :)

  7. 11
    Erin Forrester says:

    Arrived here via the thoughtful questions you posed in your LinkedIn post. https://www.linkedin.com/pulse/digital-career-9-24-240-month-plans-avinash-kaushik/

    My biggest reminder is that I am in control. Our organization like many does quarterly goals and reviews. This narrows the horizon that I as an individual bring to career discussion. It is the behavior that lulls us from the medium and long term planning encouraged in this post.

    Thanks for the list of books and people in the middle of the post. Books ordered, alerts created.

  8. 12
    Alice Cooper's Stalker says:

    Avinash,

    Great and timely article as I begin a job search. I've come to realize that for the past 3 years, I largely have been doing the same tasks with the same tools…not growing much. I just had a conversation with an old friend of mine last night where we both admitted our fear of becoming obsolete and talked about strategies to stay marketable. Your post is far more futuristic than where our discussion went.

    As for your discussion on machine learning and AI, a man that I have a great deal of respect for is James Cameron. He is uber creative and very technical. He is working on a new Terminator movie that will be a direct sequel to Terminator 2 pretending that Terminators 3 – 5 never existed. In recent interviews he makes it clear that AI and machine learning are going to be a big part of the story.

    "Technology has always scared me, and it's always seduced me. People ask me: "Will the machines ever win against humanity?" I say: "Look around in any airport or restaurant and see how many people are on their phones. The machines have already won." It's just [that] they've won in a different way. We are co-evolving with our technology. We're merging. The technology is becoming a mirror to us as we start to build humanoid robots and as we start to seriously build AGI — general intelligence — that's our equal. Some of the top scientists in artificial intelligence say that's 10 to 30 years from now. We need to get the damn movies done before that actually happens!"

    You can read James Cameron's interview on the movie and AI here – http://www.hollywoodreporter.com/features/james-cameron-sounds-alarm-artificial-intelligence-unveils-a-terminator-21st-century-1043027 .

    Thanks for giving us so much to think about and so many resources to absorb in this post. I like your "box" model.

    Alice Cooper's Stalker

    • 13

      A C S:: I agree with you on Mr. Cameron. He has always been a few steps ahead of us in having a vision for technology can do. I had not read the quote you share, but I'm amazed at how the time horizon he shares is one I believe is going to be transformative for humanity in this context.

      I have to admit one place where I had a humble difference in pov with Mr. Cameron is the humanoid robot bit. Yes, there might be a bit of that, but in the near-term we will take control of the biological interfaces with technology to become far superior versions of humans. In the long-term, I'm not sure these inefficient things called bodies will be the form we manifest (and now we are looking 100 – 200 years out).

      Avinash.

  9. 14

    Great post Avinash as always.

    My role in web analytics has changed immensely since I started 7 years ago. Back then reporting was manual, tedious and took up the bulk of my time with less emphasis on analysis – the 'analysis' was the role I was hired for. Fast forward to 2017 and reporting makes up less than 10% of my role and is fully automated into beautiful interactive dashboards.

    Now the majority of my time is spent on providing actionable insights not just on my company's website but for the app and other digital platforms. My role is now more of a digital analyst which keeps me on my toes, ever evolving which I love about working in this industry.

    Am I scared of automation? No. I've been embracing it and as long as you up-skill then I see no reason than to appreciate the new opportunities it brings.

    Olivia

  10. 15

    Rich, heavy, ever insightful as always. Thank you for sharing your thoughts. It will take some time to digest in the days ahead.

    Now, I was at a Data Science conference yesterday and one of the speakers said something like "we're living in the age of the rise of humans" alluding to the fact that automation will, contrary to the general anxiety, actually help humans develop better skills, and do better jobs.

    What do you think about this?

    • 16

      Wumi: I touched on this exact thought in my earlier post, see the Wait, so are we doomed because of AI section.

      Short summary.

      As a step one I believe that humans will connect and be infused with all kinds of intelligence, memory and compute power. We will be unimaginably better versions of humans. Your Data Science speaker was right.

      Technology (AGI, Super Intelligence, Quantum Computing…) will continue its ceaseless progress and become more intelligent and competent than all humans combined. What happens then, say 50 years from now, will be phase two.

      Avinash.

      • 17

        Thanks for replying. OK, this certainly takes off in a scary direction. 'Upgraded humans infused with tech and compute power' was clearly not on my speaker's agenda. All he meant was that AI would provide us better information and metrics to work with so we can do better, more meaningful jobs while the lesser mundane, redundant labour such as data entry would be handled conveniently by the machines.

        [Probably off-topic: I also personally struggle with imagining 'phase two' with my christian belief. Are we in denial or is this vision of ordinary human>cyborg>AI-domination a result of too much sci-fi?]

        • 18

          Wumi: "Phase Two" is a deeply personal reflection for all of us spiritual people (of traditional faiths or otherwise). It brings up tough questions, answers to some that we are usually unwilling to accept/think through.

          To your sci-fi point… To the extent my humble skills allow, I believe movies spark our imagination, but their imagination is far too limited by the anchors of the current reality. Over a long period of time, I'm optimistic about our future as a species (and I'll share my thinking on this in TMAI #100, if you are on my newsletter please check it out and share your reflection with me).

          Avinash.

  11. 19

    Avinash,

    Thank you for another forward-thinking post! I look forward to exploring the resources you've recommended to stay abreast of the movements in AI. Occam's Razor is certainly the top of my list for this!

    For me this movement seems so much more exciting than it is scary. It means automating and minimizing the soul-destroyingly dull parts of my job and focusing more on the reasons that I got into digital marketing and analytics in the first place. The things I was passionate about during my marketing courses in university; thinking critically and analytically to understand our consumers and strategy on a deep level and then applying creative solutions to take action and help the business thrive. These skills require a human (at least for the foreseeable future). I think the new focus of effort will only help humans become more valuable in their roles.

    This reminds me of your post on the ideal balance between DC | DR | DA. Not many marketers or analysts get excited about the highly repeatable DC | DR work (at least none that I know), but some do find security there. I'll admit, this was the case for me. I was unsure about automating/ minimizing DC | DR in my role. But when we did I was able to focus on great insights and my value to the organization increased 100x along with my job satisfaction. I hope that advancements in AI will bring the same kinds of transformation!

  12. 20
    Donovan Cooper says:

    Most agencies are fearful of the automation that is quickly coming to every part of Digital Marketing. Agencies that have a deeper history in television or print where here was no need to learn anything new for decades are truly freaking out.

    I share the optimism for similar reasons to the ones Justin mentiones in his comment.

    We will all have access to the same core Machine Learning algos that power automation which makes it vital to do the box thought experiment you recommend. What is our true value as an Agency? How can we extend what is at the very core of our competence into new areas? There are plenty of problems to solve.

  13. 21

    I appreciate your very useful resources to explore the future of work, when it comes to analytics and AI!

    A typical question for practical application of AI is how to deal with situations when our (limited) human mind cannot grasp very complex decisions made by the machines. In your link to the "Superintelligence" video, Bart Selman provides a very interesting rebuttal to this concern. In fact, as AI keeps advancing, machines will soon be able to generate explanations to their decisions that are … human understandable!

    I think that the role of Analysts will evolve to identify new opportunities to utilize AI to help companies not only address the issues the companies traditionally try to resolve (in the field of ecommerce: drive more traffic, increase conversion rate, etc.), but also identify solutions for the problems that currently fly under our radar (answering questions we have yet to ask.)

    • 22

      Alex: To add to your example, Google has also shared publicly that it has approaches to get an "explanation" from Deep Neural Networks how they are, for the lack of a better word, thinking.

      For a little while, while we are still in the world of narrow AI, this will work. The problem is when the complexity they manage, the scale at which they work, the size of challenges they are solving.

      Here's a metaphor, let me know if it helps (/if you agree)… Consider Squirrels and humans. A human can explain, in simple, logical, detailed terms, exactly how we approach growing carrots or designing a rocket ship. Yet, at no level can a Squirrel understand anything we explain.

      We are, at some point in the next five years, going to be the Squirrels. :)

      Avinash.
      PS: I'm not saying this is a good thing or a bad thing. I am not freaking out about it. It is what it is.

      • 23

        Thanks Avinash, your metaphor is spot on! While I don't personally think it's a great thing to be a non-contributing squirrel, I agree that I have no choice of accepting or rejecting this situation, it simply exists without my input. I do wonder if it would take AI developers only 5 years to get to the point when as humans we would be able to understand the machines.

        In the meantime, perhaps we should use your "So, what" test? Surely as squirrels we don't understand the process of growing carrots or designing a rocket ship, however, we CAN understand the implications of having carrots and rocket ships at our disposal and even possibly observe the patterns behind different harvest seasons and geographical availability of such goods.

  14. 24

    Avinash, love your posts and emails. They totally motivated me to kick off my AI and ML self-learning journey.

    I've been thinking whether 'good' data (i.e. Complete, accurate and facts-reflecting) can still be bad from an ethic-centric POV because good data capture both good and bad decisions made by human, so it contains all kinds of biases in it. The challenge is how to turn 'good' data into 'great' data (i.e. Bias-free data) if it's indeed possible….what do you think?

Add your Perspective

*