You don't have to join. Nobody is twisting your arm.

Outfuel crafts permanent digital exhibits showcasing the latest scientific discoveries. Sign up below if you're in to that kind of thing. Otherwise, enjoy the exhibit.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Already a traveler?
january 2019


Outfuel is a self-publishing network of insightful writers, thinkers, and storytellers. We shine light on topics related to the future of human-computer interaction by releasing digital experiences like this one.

The Future of


the world was awakened by the silent gold rush in AI. Giants erupted, amassing fortunes almost overnight.

Like an earthquake sparking a tsunami, the future Wolves of Wall Street joined the hunt. Yeah, so what? Well — are you going to let this moment just pass you by? Or, are you going to stop fiddling around and f*ing pay attention? Still around? Great.

In 2018,

Scroll down


Quick, Everyone talk about AI

If you look hard enough, you'll find a Tsunami of information about artificial intelligence.
We dove in, headfirst.

In 2018 alone — thousands of books, research papers, podcast episodes and TED talks tried to make sense of the growing wave of disruptive machine learning technologies sweeping the globe.

An Epic Wave of Knowledge, Technology & Change

This isn't your daddy's AI. While the previous waves of artificial intelligence technology ended up to be small swells, this wave is shaping up to be a tsunami. The world is waking up and getting out their surfboards. Open-minded pioneers are already riding this big kahuna into the future.

Applications of AI are everywhere. Designers are using AI to choose their color schemes. Companies re democratizing AI for marketers. Microsoft is using AI to combat Russian hackers. Scientists are using AI to develop electronic tongues that can taste. IBM is using AI to drive more informed debates. Machines are even starting to teach each other.

It's your choice. You can either decide to ride this wave, or get left behind.

Finally, an age-old dream is starting to crest on the horizon. In our eternal quest to become masters of our future, we’ve looked to the stars, to the oracles, and now we've turned our gaze to data. And still, we yearn for a better way to predict our own future.

In order to find a better way, the best ideas need to light the path forward. Luminaries are everywhere. But, their light is drowned out by the 24 hour news cycle and the modern drive to get more done everyday.

Some people see things that are and ask, Why?
Some people dream of things that never were and ask, Why not?
Some people have to go to work and don't have time for all that.
- George Carlin

You might think there's not enough time in the day to go through all of this information. You're almost right. We spent months curating, researching, and interviewing to produce and publish this interactive publication.

We've Gathered It All Here In This Living Publication

In our 4-part series on Prediction, we're cutting through the jargon, PR talk and fake news. We're taking our subscribers on a journey through the world of AI, machine learning, and deep learning starting with this (our first) issue entitled The Future of Prediction.

This issue is devoted to giving you a crash course on the latest in AI. Keep reading to access to exclusive content from up-and-coming prediction rockstars and collections of the best books, talks, companies and people in AI.

In the next 3 editions, we'll explore how AI is shaping:

  • The Future of Feedback & Answers
  • The Future of Questions & Decisions
  • The Future of Debate & Clarity

Want To Ride This Wave With Us?

Riding this wave alone would be lonely (and boring). We believe that intelligent debate and discourse will help shape the future. So much so that we created Outfuel, a digital place to dive headlong into the great debates of our time.

👇 Sign up to get notified when we release our next installment 👇

To follow along


or subscribe here

Jan. 2019

AI turns brain activity into speech

For many people who are paralyzed and unable to speak, signals of what they'd like to say hide in their brains. No one has been able to decipher those signals directly. But three research teams recently made progress in turning data from electrodes surgically placed on the brain into computer-generated speech. Using computational models known as neural networks, they reconstructed words and sentences that were, in some cases, intelligible to human listeners.

SEP. 2018

Can artificial intelligence help stop religious violence?

Software that mimics human society is being tested to see if it can help prevent religious violence.Researchers used artificial intelligence algorithms to simulate actions driven by sectarian divisions.Their model contains thousands of agents representing different ethnicities, races and religions.Norway and Slovakia are trialling the tech to tackle tensions that can arise when Muslim immigrants settle in historically Christian countries.The Oxford University researchers hope their system can be used to help governments respond to incidents, such as the recent London terror attacks.However, one independent expert said that the tool needed more work before it could be used in real-life situations."This could be an extremely useful research project when it reaches maturity as a thought tool for analysing factors involved in religious conflict," said Prof Noel Sharkey.

SEP. 2018

AI Will Create $13 Trillion in Value by 2030

Consultant firm McKinsey reckons artificial intelligence (AI) will boost productivity way more than the steam engine was able to in the 1800s.McKinsey's latest forecast of AI's impact on the global economy is that it will have generated $13 trillion in economic activity across the world by 2030, despite causing upheaval for many people.

Jan. 2019

Recruiting AI Talent: 4 Ways To Get Ahead Of The Next Hiring Wave

Usually I talk about AI and how it’s already changing recruiting. Or I did about five minutes ago. But let’s talk about something else: AI jobs themselves are going to change recruiting as well — by putting increasing pressure on recruiters to fill the soaring number of AI-related positions.The World Economic Forum estimates that by 2022 we’re going to create 133 million jobs in artificial intelligence and machine learning. Spending on AI is projected to surge — from $12 million in 2017 to $57.6 billion by 2021. And a recent survey found that 61% of organizations most frequently picked Machine Learning / Artificial Intelligence as their company’s most significant data initiative for next year.

Jan. 2019

AI, Machine Learning Jobs Among Highest Paid in Tech

Artificial-intelligence and machine-learning professionals are expected to earn some of the highest salaries for information-technology jobs in the year ahead. AI developers and machine-learning engineers are garnering average annual wages of up to $200,000, according to a report from New York staffing firm Mondo on Wednesday.  Those developers and engineers last year earned $150,000 and $175,000, respectively, the company said.

Sep. 2018

Would You Welcome A Robot Boss? This Study Thinks So

Perhaps it’s the proliferation of ‘soft’ robotics— everything from the BB8 droid in the Star Wars movies to Pepper, the world’s first humanoid robot. Whatever the reason, Brits are currently loving their artificial intelligence, so much so that a new study has found that 53% of employed workers would be happy to work for a robot.

Jan. 2019

Hiring For The AI (Artificial Intelligence) Revolution - Part I

In the coming years, Artificial Intelligence (AI) is likely to be strategic for a myriad of industries. But there is a major challenge: recruiting. Simply put, it can be extremely tough to identify the right people who can leverage the technology (even worse, there is a fierce war for AI talent in Silicon Valley).To be successful, it’s essential for companies to understand what are the key skillsets required (and yes, they are evolving).

Apr. 2018

How AI Can Save Our Humanity

AI is massively transforming our world, but there's one thing it cannot do: love. In a visionary talk, computer scientist Kai-Fu Lee details how the US and China are driving a deep learning revolution -- and shares a blueprint for how humans can thrive in the age of AI by harnessing compassion and creativity. "AI is serendipity," Lee says. "It is here to liberate us from routine jobs, and it is here to remind us what it is that makes us human."

Learn the


Artificial Intelligence

Used in a sentence: Using artificial intelligence, the app predicted when someone was likely to purchase.

You know this one already. AI is the field of computer science that seeks to enable machines to do tasks that would normally require human intelligence.

flip →

Further Reading

If you'd like to learn more about artificial intelligence you can check out:

HackerNoon— An Introduction​ to Artificial Intelligence
Geeks for Geeks — Artificial Intelligence | An Introduction
Medium — Introduction to Artificial Intelligence



Used in a sentence: Netflix uses algorithms to generate your recommended television shows and movies.

An algorithm is a formula that represents the relations among a set of variables. As it relates to AI, these are what machine learning programs use to make predictions from data sets. You see algorithms in action every time you sit down to watch Netflix. All of the content suggested for you is based on algorithms that have been written to learn about your movie watching preferences.

flip →

Machine Learning

Used in a sentence: The supervised machine learning algorithm needed more training data.

This subdiscipline of AI is just what it sounds like: The ability for computers to learn without being explicitly programmed. In practice, this could manifest as discovering new patterns in datasets, face recognition, or the creation of predictive algorithms. There’s supervised learning, in which humans need to input data and, as the name implies, supervise the process; and unsupervised learning, where the program is essentially left to find patterns and draw its own conclusions.

flip →

Further Reading

If you'd like to learn more about machine learning you can check out:

Stanford School of Computer Science — Introduction to Machine Learning
Geeks for Geeks— An Introduction to Machine Learning
Medium Article — Introduction to Machine Learning — An Introduction to Machine Learning Algorithms


Natural Language Processing (NLP)

Used in a sentence: Siri uses NLP to make herself construct sentences like a human.

This is the computer's attempt to understand spoken or written language. (And it often involves machine learning.) You interact with it every time you ask Siri or Alexa for directions to the nearest Starbucks.

flip →

Further Reading

If you'd like to learn more about natural language processing (NLP) you can check out:

Medium Article — An Easy Introduction to Natural Language Processing
SAS — What Is Natural Language Processing
Forbes — What Is Natural Language Processing And What Is It Used For?
Medium Article — Natural Language Processing is Fun


Neural Network

Used in a sentence: Our software uses neural networks for time series predictions.

If AI seeks to replicate human intelligence, then the neural network is the brain of the operation. It’s highly abstract and simplified, but it’s modeled on the layered operation of neurons in the human brain.

flip →

Further Reading

If you'd like to learn more about neural networks you can check out:

MIT — Explained: Neural networks
Skymind — A Beginner's Guide to Neural Networks and Deep Learning
Techcrunch — Neural networks made easy
The Nature of Code — Neural Networks


Deep Learning

Used in a sentence: The self driving car used deep learning to identify a stop sign.

Think of deep learning as the way neural networks are able to see incredibly complex patterns in data sets. It allows a neural network to learn on its own by loosely defining parameters, allowing the computer to begin recognizing patterns.

flip →

Further Reading

If you'd like to learn more about you can check out:

SAS — What Is Deep Learning?
Skymind — A Beginner's Guide to Neural Networks and Deep Learning
Mathworks — What is Deep Learning?


What should We Add To This List?

Thank You

Your Content Will Be Added Soon

Our team's checking over the info you submitted. It'll be added to the repository soon!
Oops! Something went wrong while submitting the form.

Prediction 101

Fundamentals & Terminology

transformative thinking

Originals chosen by our great community.

keep reading

from the founder of clearbrain, Bilal Mahmood

How to Turn Machine Learning Into Action With Predictive Personalization

Harnessing the Power of Prediction

Every marketer wants to be as good as Netflix or Amazon at predictive personalization.

These best-in-class companies are able to deliver hyper-personalized experiences to their individual users, from serving the right product category to a buyer at their moment of highest intent, to offering the right discount rate to a customer proportional to their risk of churn. In turn, such predictive personalization enables these companies to drive incremental revenue by marketing to the right person, at the right time, at the right place.

Delivering such experiences requires a data and marketing infrastructure established from the ground up for AI automation. Without AI, a marketer has to guess at the rules for segmentation in personalizing experiences - they’ll configure their audiences by a couple basic demographic or behavioral rules to construct groups of users they believe share the same intent. With AI however, one can truly personalize a user experience by drawing from hundreds, if not thousands, of user signals and transforming them into a single propensity score identifying each individual user's likelihood to perform a specific action within a specified time and channel.

Given advances in marketing and data infrastructure, it’s now possible for any company to deliver such 1:1 user experiences. Any marketer can now leverage a sequence of tools to identify their high potential users, construct the right message for them, and deliver it at the moment of highest intent.

In this, article, I outline how you can construct your own predictive personalization stack in three steps:

  1. Data Instrumentation in which you map the digital footprint of your users across their customer journey
  2. AI-Based Segmentation, where you convert those digital signals into intent-based user segments
  3. Predictive Personalization, where you translate user intent into recommended actions to drive incremental gains.

Step 1: Data Instrumentation

The first step in building a predictive personalization stack is data.

Building predictive experiences is all about constructing a persona of an individual user to estimate their future intent based on past intent. In turn, in order to measure and build this persona based on past intent, you need to map your user’s past behavior into a digital footprint that can be analyzed. The classic mantra of “garbage in, garbage out” very much applies here.

So the first step to enable a predictive personalization experience is to instrument your app to log every user attribute and action, from first visit, to sign up, to targeted conversion goal. Customer Data Platforms like Segment and mParticle are perfect for this - they serve as a universal user tracking layer to map your user data in a format conducive to downstream analyses.

In using these tools, it’s important to instrument and log your app to track the breadth of all your user attributes and actions. First, start by mapping your user journey. Construct a step-by-step outline of the typical user path to conversion, and then make sure each step is logged and recorded.

There are multiple ways to record these steps.  Within tools like Segment and mParticle, you can log each step as a pageview, screen view, session, or custom event. Make sure to choose at least one of them to have a digital record of the action having occurred.

User attributes is an additional record to keep. This include attributes about a user that are not time-sensitive; i.e., referrer source, browser type, domain, UTM, etc. This provides important demographic information about a user to build the user persona you are seeking.

Last, it’s important to map the user journey across multiple domains or assets you may have. Don’t  instrument your main app only, but also your marketing site, landing pages, blogs, community pages, etc. Each individual domain provides an important step of your user’s journey, and their behavior in each domain may provide unexpected clues as to their intent in different assets in step two.  

Step 2: AI-Based Segmentation

With your user journey mapped, you have the data infrastructure necessary to build AI-based segments.

Now remember, the goal of an entire predictive personalization framework is to optimize the user experience for a specific conversion event, say, purchase.

With a traditional segmentation approach, you’d build an audience based on a set of rules or stages of the customer funnel you think are most indicative of intent to purchase. For instance, you may construct a purchase audience based on users who did an add-to-cart action last week, and purchased at least once before. This is likely sufficient, and helps with a theoretical 2% conversion rate. But how about moving the conversion rate from 2% → 5%? This would require choosing more and more rules to segment an audience, which may not be intuitive; i.e., the user visited the add-to-cart page 5X last week, or the user has never purchased, but their engagement rate increased 10% WoW for the last four weeks.

Capturing the full breadth of such signals requires a more automated approach. We need to flip traditional segmentation on its head. Instead of segmenting users by 2-5 demographic and behavioral attributes that you instrumented in Step 1 (which you think lead to purchase), you instead segment by a single number - their intent to purchase.

Segmenting by intent is only possible through artificial intelligence. Using an AI or ML based approach is simple in theory. You look at which users performed a purchase action in the past, compare them to users who did not purchase in that same period, and assign relative weights to all the user attributes you instrumented in Step 1 with higher weights for those attributes that help distinguish between users who did vs did not purchase.

An AI-based segment in turn is able to combine all your instrumented demographic and behavioral attributes, determine just the right weight for each one, and give you a single probabilistic score to segment users by high vs low intent.

Now the math behind constructing this AI-Based segment are a little complicated. It involves using machine learning approaches, from regression to deep learning techniques, that normally require an army of data engineers to transform your data from Step 1 into a machine-readable format, and then data scientists to construct the intent-based segments. However, there are new predictive analytics platforms, like ClearBrain, that enable you to automate this entire process down to minutes.

With predictive analytics tools, you can ingest your user data mapped in Segment or mParticle, and the ClearBrain platform will automatically weigh every user attribute and event and give composite probabilistic scores for any target behavior.

This enables you to construct intent based audiences - point-and-click. Simply select the target behavior you’re seeking to optimize, and the platform will assess and weigh every other user action and attribute and its relative importance in differentiating between users who will vs will not do the desired outcome.

Step 3: Predictive Personalization

Once you’ve built your AI-based segments, you’ll have successfully built audiences by their intent to perform a specific action.

But simply knowing that a user is most likely to purchase doesn’t tell you how to actually influence the outcome of their action. Merely understanding a user’s intent isn’t all  that helpful, as it’s telling you their likelihood to perform the desired action in the absence of any marketing action.  What you really need is a way to leverage the user’s propensity for an action as a foundational input into driving a customized experience.

There are three primary ways of personalizing an experience based on a user’s intent: targeting, content, and pricing.

Method A: Personalized Targeting

The simplest way to leverage AI or predictive-based segments is by knowing  who to target in your marketing campaigns at all. Here, there are two strategies, depending on whether you are optimizing for conversions or for efficiency.

If optimizing for maximum conversions in the short-term, having predictive-based segments allows you to group your users into high vs low intent to purchase. In turn, you can exclude all low-intent users from your marketing campaigns, and just focus on the users in the medium to high-intent. This will enable you to focus your ads, emails, and push content on people for whom there is a likelihood to actually perform the action you are desiring.

If optimizing for efficiency over the long-term, however, you can use predictive-based segments to instead focus solely on your most incremental users (users who would not convert without a touchpoint). In this scenario, you’d actually exclude both the lowest intent and very highest intent users from your marketing campaigns - because the highest-intent users are going to convert with or without your ad $s, and so you are wasting CPA on those users.

Whether optimizing for conversions or efficiency, however, predictive segmentation provides a foundation for an AI-based approach as to which users to target in your marketing channels.

Method B: Personalized Content

With targeting conditions established for a marketing campaign, the next optimization you can make to your campaign is to personalize the content.

AI-Based Segmentation enables you to group your users not just by their intent to purchase a product, but also by the type of product. With a rules-based approach, you’d be able to generically group users roughly by their likelihood to purchase a product vs not. But you would not be able to determine which users are interested in Product A vs B vs C. This is what companies like Amazon or Netflix are infamously good at - extrapolating from thousands of digital signals of users to build intent-based audiences with an affinity for  each individual product.

Personalized content comes into play in different ways depending on your industry. If you’re in a media company, you can use AI-based segments to personalize categories of content and then serve them live via tools like Optimizely;  i.e., grouping users into high affinity audiences by interest in entertainment vs. sports vs politics. By contrast, if you run an eCommerce shop, you could personalize your ad retargeting to show users images of fruit vs vegetables vs dairy depending on which predictive segment they fall into that day of the week. And if you’re a SaaS business, you can recommend different subscription plan levels when a user is ready to upgrade, sending a personalized email to a user indicating that they should subscribe to self-serve vs an enterprise-grade plan.

By recommending the right type of content, the benefit is one of maximizing conversions. If you recommend the wrong content or message, you risk your message being perceived as irrelevant, or at worst hurting your brand. If you push someone to an enterprise plan when they in fact only have the budget for a self-serve plan, you lose out on potential revenue. Being able to recommend the right message to the right user enables a holistically better experience for both users and marketers.

Method C: Personalized Pricing

The last lever at your disposal  in enabling a personalized user experience is one of price.

If you know a user’s intent to purchase, the impressive facet of that knowledge is that you can customize your product monetization strategy to map to each user’s price elasticity of demand, rather than a one-size fits all model. Rather than charging each user the same price for the same subscription plan, what if you could offer different plan types based on their affinity to buy?

This strategy is most common in ad bidding, and actually inherent to how ad exchanges set pricing in their platforms. If you know your user’s intent to purchase, you can in turn personalize the bid price you input for them in the respective ad network by their actual likelihood to purchase. Bid lower for low-intent and high-intent users (because their likelihood to purchase is unlikely to change), but instead bid higher for your medium intent users.

Translating this strategy over to subscription discounts, there are comparable strategies. When a customer is up for subscription renewal , it’s common for media companies to offer a blanket 20% discount to convince someone to renew. But if you think about it, there is a significant percentage of users with high intent who would have renewed without a discount - resulting in a high opportunity cost and lost revenue. With a personalized approach, you could instead customize the discount to a rate proportional to a user’s likelihood to churn. Uber is famous for doing this to the extreme level - where they will customize rider fares to each individual ride and person, so much so that two people going from the same origin to the same destination may in fact see different rates!


What if you could predict the future of innovation?

Well, as it turns out, you can — sort of. Vittorio Loreto, professor of physics of complex systems at Sapienza University in Rome, director of the Sony Computer Science Lab in Paris, and faculty at the Complexity Science Hub in Vienna has (along with colleagues) created the first mathematical model that accurately reproduces the patterns of innovation. According to the MIT Technology Review, “The work opens the way to a new approach to the study of innovation, of what is possible and how this follows from what already exists.”

We spoke with Professor Loreto to find out more about his theories, as well as the lessons we can learn and actions we can take when we’re searching for new, innovative ideas of our own.

Want to learn more? Click the button below👇

Watch his Ted talk

Or, for some more reading

keep reading

from The Founder of Buyer Personas, tony Zambito

Are Buyer Personas Relevant Predictors In An Artificial Intelligence World?

The growth in Artificial Intelligence has been astounding over the past three years.

The business world has seen an explosion of AI, as everyone from start-ups to major entities like Microsoft, IBM, and Amazon make AI a significant offering.

It appears we are in another cycle of technology disruption. In the 1980’s it was the arrival of personal computers, ushered in by a glut of computer and hard-drive makers, with only a few still surviving. In the early 2000’s we saw the explosion of the Internet and the boom ... followed by the bust. However, the Internet has connected every corner of the world and profoundly changed the world as we know it.

The advent of Artificial Intelligence has resulted in another profound shift that’s currently underway. AI will significantly alter how humans interact in a digital-centric world. In some ways that’s good; and, in other ways, it’s sparking controversy over privacy. While questions regarding privacy concerns remain, businesses and commerce today are under enormous pressure to adapt, as they acknowledge the risks involved in not doing so.

As a result of Artificial Intelligence, questions are being raised about the relevancy of buyer personas.

Many of the doubt-filled questions are coming from the Artificial Intelligence start-ups themselves, who maintain that AI offers the ability to develop buyer personas that are data-driven, thus, creating more accurate and real-time buyer personas.

As is the case with most business proclamations, there are kernels of truth here. At the same time, there are many proclamations that are ridden with false assumptions, which is the case when discussing the relevancy of buyer personas and the advent of Artificial Intelligence.

It is important to first understand the original intent of why personas came to be.

After the disruption of personal computers in the 1980’s and 1990’s, the world was fraught with poor product and software design. A concept was born: If developers can understand, at a deep qualitative level, the goals that influence usage behaviors, then the results can be user-friendly products. This same line of thinking was then applied to innovating buyer personas, primarily, with the intention of seeking to understand the goals influencing buying decisions.

At the heart of buyer personas is this core truth regarding goals. The body of research in the social sciences regarding goals is one of the main influences on the invention of personas. This body of research has predominantly found that the pursuit of choices and decisions are largely goal-directed.

It is through this deep understanding of goal-directed behaviors that marketers can begin to unveil the answer to the all-important question: Why do users and buyers make the choices and decisions they do? To answer this question, the best proven means has been the use of qualitative research. It is a question that Big Data and Artificial Intelligence alone cannot answer.

Where Artificial Intelligence can be helpful is in monitoring the pursuit of goals accomplishment.

Artificial Intelligence enables the monitoring of clicks and intent, providing more real-time discovery of clickable behavior patterns, which offers insights into the where, when, and how of the pursuit of goals. For example, a simple search for a used car can lead to Artificial Intelligence serving up cars that appear to match a person’s online searching. However, truly understanding why a person or group of people are pursuing a used car cannot be ascertained from Artificial Intelligence alone.

User and Buyer Personas were meant to provide a human face for understanding users and buyers, helping us to humanize our understanding of who they are and what they are attempting to accomplish. In a digital-centric world where we are encountering Artificial Intelligence, this need to humanize our understanding remains essential to commerce. Perhaps, it’s even more important as consumers and buyers further retrench behind a digital veil -- the backlash against too much invasion of privacy.

Are buyer personas still relevant predictors in an Artificial Intelligence world? The case can be made they are very much so. Buyer personas tell us a lot about what goals will drive buyers in the future, providing insights into what buyers aim to accomplish and why. Furthermore, they can serve as a good predictor into how they will pursue goals. Such a degree of data can then make Artificial Intelligence even more beneficial.

Smart organizations in the future will help their customers with both personas and the use of AI. Each can inform the other and offer rich insights. The smart approach is not to forgo one for the sake of the other. Instead, utilize both in concert to truly understand customers in a rapidly-evolving digital-centric world of commerce.

keep reading

from AUTHOR, Victor Antonio

How Is AI Changing the World of Sales?

Victor Antonio is a Georgia-based sales trainer, consultant, and motivational speaker with an international client base including Aflac, Fidelity, and Cigna.

He’s also the author of 13 books about sales and motivation, most recently Sales Ex Machina: How Artificial Intelligence is Changing the World of Selling. So naturally, we just had to pick his brain for the Prediction Issue.

Victor was kind enough to share his thoughts on AI-fueled sales, the importance of fostering a data-driven culture, displacement, and more with Outfuel editor Trix Middlekauff.

Our Q&A With Victor

Can you briefly talk about your background in terms of how you came to be a proponent of the application(s) of AI in sales? What experiences led you there, and what personal successes have you had with it? What inspired you to write Sales Ex Machina?

In your experience, what are the worst fears that sales & marketing team members have when it comes to AI and machine learning?

Along the same lines, what are the biggest misconceptions people have regarding the use of AI to assist/optimize the sales & marketing process?

How can AI-fueled (would partnered be a better word?) sales help improve the customer journey beyond the point-of-sale? (As in properly setting expectations, fueling engagement, etc.?)

In your book, you talk about the creepy “deja knew” phenomenon. .So … how can sales & marketing leverage the data-gathering power of AI while also circumventing that unnerving feeling of “How did you know so much about me?” In other words, how can you straddle that fine line between helpful data- fueled action vs. creeping people out, which can be counter productive?

As I understand it, you’re saying that NLP and Sentiment analysis is a way to sort through the customer feedback that’s “out there” and take the right action(s)? Can you address the strengths and weaknesses of this process? (For example, you point out NLP’s issues with understanding or correctly interpreting sarcasm.)

In the book, you write: “A data-driven culture is one that respects the process and rationale behind the data being collected, and of course, the actionable insights generated by the machine. Gut instincts and trends take second and third compared to what the data is indicating. ‘Shooting from the hip,’ as many are wont to do, is a serious “No-No” in the data-driven milieu. This is a difficult culture shift for many, particularly those traditionalists harking back to a bygone era.” That said, do you make a distinction between being data driven vs data informed? If not, why not, and if so, are there any inherent dangers in prioritizing one over the other? How can one avoid those pitfalls?

Let’s say I’m in the sales and marketing department at my company. The higher-ups are, in a word, Luddites. I really want to learn about and leverage the power of AI to help me be better at my job. But they don’t get it; they won’t cooperate. What can I do on my own? Any advice?

In the book, you maintain that companies need a collaborative AI strategy to remain competitive. Does this apply to all sizes and types of companies equally? I.e., startups vs. enterprise vs B2C vs B2B etc.? What unique challenges or impediments may apply to different business models?

For those who haven’t read the book ... can you explain the meaning of “the singularity” and address & contextualize the fears that many seem to have that the All Powerful Machine will eventually replace (if not destroy!) them? How can sales teams stay ahead of replacement, and be prepared for displacement? Should some people start thinking about a different area of sales, for example?

Read More

Did Victor’s insights in the interview whet your appetite to learn more about sales & AI? Then get the Sales Ex Machina book and read more!


If it has anything to do with upping your sales and marketing game, chances are Victor Antonio has made a video about it.

Check out his lively and engaging presentation style on his YouTube channel, his live speaking events, and his online training academy. And, if motivational public speaking is your thing, you’ll definitely want to check out the award-winning documentary “The Motivator: The Business of Selling Hope” and the recent film “Beyond the Stage,”  available on Amazon Prime.

keep reading

from The Founder of Obviously AI, Nirman Dave

Spotlight on: Obviously AI

“It can’t tell you everything about the world, but it will tell you everything about your data.”

For most undergrads, just making it to a Monday morning class on time is a major Accomplishment. Twenty-year-old Nirman Dave, however, is not your typical undergrad. Now in his senior year at Hampshire College, Dave doesn’t ever have to wake up in the wee hours to make it to class. In fact, he never goes to class at all. Instead of worrying about term papers, grades, and lectures, Dave’s entire senior year is focused on launching a Silicon Valley startup, Obviously AI.

“You could say it’s the Google for enterprise data,” Nirman says of his AI-driven brainchild. In a nutshell, it works like this: Product analysts upload data sets, choose custom settings, and input natural language queries about customer behavior like “Who will churn?” or “Who will upgrade?” . Then, they receive “instant predictions in minutes” with, according to Dave, 87 to 92 percent accuracy. No data science, machine learning, or math chops required.

That promise of accurate predictions within minutes piqued our curiosity, and so Outfuel chatted with Nirman to find out more about Obviously AI and his vision for his fledgling company, which he co-founded with Tapojit Debnath. Here's what we learned.

What drives Nirman Dave?

“I get a lot out of [studying,] Nirman told Outfuel. “I've been studying machine learning and behavioral economics, and I love that. But I really love the idea of action and being around people and working with people to get somewhere. That drives me everyday. So that's the kind of mindset that I’m coming into it with right now.”

What was the inspiration for Obviously AI?

There’s no doubt that Nirman Dave was born with an inquisitive and entrepreneurial spirit. He taught himself how to code in the 6th grade, developed his first app in 7th grade, and was teaching a college class in India by the age of 17. In his freshman year at Hampshire, he organized a wildly successful Hackathon that paired coders with people of other disciplines, resulting in, among other things,  EdgarAllanPoetry, an AI that “tests users’ ability to tell computer-generated poetry from that of the English language’s best poets,” according to a 2016 New York Observer article.

He traces his inspiration for Obviously AI in particular back to his interest in the intersection of behavioral economics and machine learning. “I took a lot of behavioral econ courses and finance courses [in school], and that really got me excited into how people think about money and psychology,” he explains. “Fast forward to my senior year. Now I wanted to do something that really combines machine learning and human behavioral economics and the sweetest kind of intersection between the 2 fields. I actually didn't plan to do [Obviously AI]. I stumbled upon a problem at a company that I interned with last summer, where the product manager didn't have a tool that was readily available to answer all their customer-related questions. I thought about how about they could Google their customer behavior as if they would be Googling about anything else in the world. And that's kinda where this Obviously AI idea came into the picture.”

So, how does this machine learning prediction thing work exactly?

The good news is, you don’t have to be an expert in machine learning or data science to use the product. But if you want to get an idea of how it works, Nirman broke it down in layman’s terms for us:

“If a manager or growth hack analyst is looking to find insights about that customer -- they have existing customer data. So what they do is the come come to our platform to connect the database that they already have with data about existing customers .. And once they connect the database they get a Google-like search box where they can ask questions like: “Who is going to churn?” or “Why did people churn in the last 6 months?” Or they can ask more deep questions like “What is common between my best customers?” and stuff like that. So they can ask all these deeper questions about customer behaviors in plain English and get answers instantly.

Now the way we get these answers instantly is, we actually take the data and throw it into a neural network that we've created … that is tailored for the company's data to understand and learn patterns and customer behavior. If you're asking questions about churn, first of all the natural language processing in the website understands what you're looking for, finds that data and throws it into an automated ML which creates a tailored neural network for the data. Then that neural network really does the classification.”

Watch the Obviously AI Demo below

How can the machine learning model account for variables that aren’t in the data?

The short answer is -- it can’t, at least not yet. As Nirman explains, “When a product manager or growth hack analyst is using our product, they use it knowing that the machine learning model will not see externalities and that is where their job comes in where they say: ‘Okay. Here's what the machine is sending me. Here's what I think could also happen.’”

That said, Nirman says he’s on it: “I'm really trying to tackle this problem by creating what we call externality scoring. That is a separate algorithm that’s still under production … [we would] look into Twitter, look into Facebook, look into a lot of different data sources specific to your company and your industry … then every prediction will give you a little thing on the side that says 'Here are possible external factors that could affect the prediction.'”

Bottom line? “It can’t tell you everything about the world, but it will tell you everything about your data.”

What makes Obviously AI different from other machine learning tools?

In a nutshell, according to Nirman the big differentiator is the combination of ease-of-use for non techies combined with hefty prediction power. “A lot of tools out there [like DataRobot] … require you to have some form of either programming knowledge or machine learning knowledge or data science knowledge,” he says. One of the biggest differentiators for us is that we don't require a product manager to know anything at all about machine learning, programming, or any kind of data science or math. They literally have to just Google a question.”

But what about a company like Thought Spot, which is also geared towards those without technical know-how? "Thought Spot, as we understand it, is a search tool that uses NLP to find existing information across different data sources,” Nirman explains. “We are a predictive analytics tool that uses NLP to predict future customer behavior using existing information across different data sources. So, we not only search the right data but also run ML algorithms like kd-trees and Neural Nets on that data to predict how customers will behave in the future. “

What does the future hold for Nirman’s prediction machine?

“The product vision is very simple,” Nirman says. “We know that every company in the world would want to use something like this ... everyone would want to have a way to Google their customer behavior  .. and not spend  hours or days of work just to get a simple answer  .. So we know that every company wants this and our target as a product is to make sure that every company in the world can use this product and will use this product.”

As for his company vision, Nirman dreams of a world where everyone is a master of their own algorithm. “Back in the day computers were very much like this, where it was only operated by scientists and researchers. So we kind of want to go that route and make machine learning personal and accessible for everyone. … Today we are doing it for product managers and non-technical people in enterprise … Moving forward in the next 5 years, we want to make it personal and accessible for pretty much everyone in the world.”

Obviously AI Launches February of 2019

keep reading

from Dr. A.K. PRADEEP

What if you could predict the future of innovation?

The field of neuromarketing purports to reveal the unconscious desires of consumers by studying the brain's response to marketing stimuli.

Author, entrepreneur, scientist, & UC-Berkeley-trained engineer Dr. A.K. Pradeep was something of a pioneer in this field. Over a decade ago. he founded NeuroFocus, a multinational neuromarketing company that was acquired by Nielsen in 2011.

Flash forward to now: Dr. Pradeep is heading up a new venture, Machine Vantage, which he founded in 2016. Among other things, his company seeks to harness the power of machine learning and AI to uncover patterns that can reveal consumers’ unconscious motivators and desires.

I’m Trix Middlekauff, editor at Outfuel, and I had the opportunity to speak with Dr. Pradeep about his work, his new book, AI for Marketing & Product Innovation, and the ways in which intelligent machines can potentially suss out the metaphors that give meaning to our world and the products we crave.

Watch Dr. Pradeep's talk on the new frontiers of neuroscience

For some more reading

keep reading

Showcasing some of the top minds in AI

Showcasing the top minds in AI

AI Photo Credit —

This image was created in part using the artificial intelligence in Adobe SenseiLuminar 3.

We are building a continuously updated showcase of the top minds in machine learning from around the world. 🎉

We realize this may be overzealous. A complete list of great thinkers in AI is a fine idea in theory. In practice, it has proven impossible to prioritize this list and confirm each short bio via email. So, we're doing the next best thing and taking our time to build out this list throughout 2019 and 2020. Follow along and we'll do the best we can to reach out, confirm the facts, and add people to this showcase.

To follow along


or subscribe here


It was the summer of1956. They had the entire top floor of the Dartmouth Math Department to themselves. Invited by John McCarthy, around 20 individuals would pass through the main math classroom. The Dartmouth Summer Research Project on Artificial Intelligence (the name of a this now-infamous, informal summer workshop) is widely considered to be the seminal event for the artificial intelligence field. You can read the full story told by Grace Solomonoff in her warmhearted 28 page biographical history.

Ray Solomonoff


It was the summer of1956. They had the entire top floor of the Dartmouth Math Department to themselves. Invited by Ray Solomonoff, around 20 individuals would pass through the main math classroom. The Dartmouth Summer Research Project on Artificial Intelligence (the name of a this now-infamous, informal summer workshop) is widely considered to be the seminal event for the artificial intelligence field. You can read the full story told by Grace Solomonoff in her warmhearted 28 page biographical history.

Gohan Lee

CEO of Muma Template

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique.

Philip Brower

UI/UX Designer at Muma Template

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique.

Sara Doe

Developer at Muma Template

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique.

Latest News

From Blog

Sorry folks. this site is meant to be experienced on a larger screen.

For now, this publication is best experienced on a desktop computer or tablet with the display width of at least 1024px.

At the moment, the best we can offer is the handy dandy reminder form below. You can use it to remind yourself to come visit this website on another device.

Get a polite reminder email sent to your email inbox.

One email. Not two. Not three. One.

No, you will not be subscribed to something.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.