Cutting Through the AI Hype

Cutting Through the AI Hype
For most of us, when the words Artificial Intelligence are in the story headline, news or in a movie, there is the image of some magical silicon brain that performs incredible feats of intellect. Problems solved and it is connected to everything so it can offer all kinds of sage advice. The reality is something quite different. I’ve done over 15 briefings on AI to C-suite executives this year alone. So what is the state of AI today really?
The Artificial Intelligence Industry is a Hot Mess
The sector of the ICT industry that is focused on AI is really quite a mess. And that’s exactly where it should be right now. There are hundreds of startups and multinationals (i.e. IBM, Google, Microsoft) working on various approaches and solutions. Consumers may see IBM Watson, Siri and Alexa at the forefront, and they are good products for sure, but they are the ones with big marketing budgets behind them too. Investment into AI is across almost every industry from retail and banking to insurance and the energy industries. All with different approaches, viewpoints and technologies.-
AI is Nascent. So Are The Technologies it Relies On
As an industry sector, AI is fairly new, only really popping into the forefront in 2011 when heavy investments and venture capital started to get involved. The technologies that Artificial Intelligence relies on are the Cloud, Big Data, Internet infrastructure and data sciences. All of which are still fairly nascent.
AI is comprised of a number of sub-disciplines such as Natural Language Processing (NLP), machine learning, data mining, predictive analytics and several others. Most of which are nascent.
There is No One Big AI Brain
Although it makes for great science fiction, there is no one big AI brain tucked away in a dark basement with awesome mood lighting. There are some very big computing rooms with massive air conditioning and energy usage though. But right now, companies like IBM, Microsoft, Apple, Facebook, Google and so on are very proprietary and very secretive for competitive reasons. Very few AI systems connect with other AI systems.
Artificial Intelligence Is Very Single Task Oriented
Just like men, AI is terrible at multi-tasking right now and for the next several years. Siri and Alexa are quite powerful, IBM Watson is amazing (disclosure; I use IBM Watson regularly for text analytics research projects and love it.) But when someone says to you “we have really cool AI in our product” they probably do, but it’s most likely only really good at doing one or two things like ordering a product or completing an information task. That’s it.
So What About The Big Brain?
For the next while, AI is going to be a hot mess. A very fractured industry with a lot of competing ideas. Some of the AI tools out there are really very, very good and they’re doing some truly amazing stuff with AI. Over the next few years, some jobs will be supplanted by AI and some cool products and companies will come along. But it’s going to be a while yet before theres any Big AI Brain out there.
What do you think?

Is Our Lizard Brain Technology’s Biggest Challenge?

Is Our Lizard Brain Technology’s Biggest Challenge?
Is it our basic survival instincts that are impacting our adoption of technology today? In the beginning, for hundreds of thousands of years, the technologies we developed were those of raw survival; tools to hunt for food, tools to prepare foods, clothes and shelter to develop. Language and sociocultural structures to establish rules of survival through cooperation. But this took a very, very long time. Technologies evolved slowly until the last hundred years. Now they’ve ramped up at an incredible pace. A rate we’ve never before witnessed as a species.
Advances in communications technologies such as the internet and the devices that enable the transfer and assembling of data into meaning (smartphones, computers etc.) mean we can share learned knowledge on a scale unlike ever before. Societies evolved because we learned how to communicate. It is this ability that lies at the very base of our social and cultural structures.
Technology Can’t Be Absorbed Fast Enough
Today, we are developing technologies so fast that I believe our species is struggling to integrate them into our social, cultural and economic systems. Perhaps it is our very basic sense of survival that tends to subconsciously reject or hold off on adopting new technologies.
We know in business that implementing new technologies in a company can be disastrous if not done with careful consideration of processes and how people work.
Case In Point: Social Media
Modern communications technologies enabled the rise of social media tools. Now, a decade later many businesses, governments and individuals are still struggling to understand what social media are, if they’re effective and why people even use them. Yet we do. If you’re reading this post, I’d wager that you know someone who thinks social media are a waste of time and useless. There were those that thought that of the printing press a few hundred years ago. And the telephone. And the television…and so on.
So What Does This Mean for Tech Companies, Society and the Economy?
Advances are so rapid that I think many societies, cultures and countries are struggling to find out how they are fitting into our social lives, business and economies. While some technologies like smartphones, laptops and tablets have reached a plateau, others like artificial intelligence and robots are advancing rapidly and they will impact our global economy. Home automation, self-driving cars, these too are going to have an impact.
I believe it is these sociocultural tensions with technology that are the biggest challenges for adoption of new technologies within society and business. Our “lizard brain” doesn’t see a lot of these technologies as crucial to our very survival – yet. Perhaps once we collectively, subconsciously, agree to the benefits, we will adopt them faster. But how to get there? The technology may be evolving rapidly, but are our social structures?
What do you think?

The Real Reason Ad Agencies Are Struggling

The Real Reason Ad Agencies Are Struggling
All marketing is about information creation and management. And this is why agencies are fading away.
Agencies Used to Manage The Talent
Designers, copywriters, ad buyers, art directors, creative directors and so on…they were all under one “roof” for decades. It was the agency management that could corral and manage them and the projects. This business model made sense, at the time. Producing creative was time-consuming and resource intensive.
Along Came Digital
Agencies first started to see fragmentation in the late 90’s as the Internet became popular, as momentum grew and analog channels became less influential, agencies struggled to maintain their model.
Then computing changed. PC’s became more accessible to people. Very little skill was needed to participate by the consumer as software improved. Then came social media, followed by smartphones and tablets. Disintermediation was running rampant. Things got even messier for agencies.
As broadband reached critical mass, so WiFi hit the world. Laptops became ever more powerful and creative software better and cheaper. This impacted all industries. Publications like Fast Company were touting the “Free Agent Nation” and they were right, but that was the early 00’s and we weren’t quite there. Now, we are and this is what is destroying the agency model.
The Rise of The Gig Economy
Around 2012, we started to see the rise of what is being called the Gig Economy, in other words, freelancers, to a degree we hadn’t seen before. The creative/agency sector has witnessed this perhaps more than any other sector.
A Shifting in Marketing Management
Companies were quick to note this as well. While there are still the majority of major brands that use agencies, even that model has shifted. Every year, less and less use AOR arrangements and more hire boutique agencies and increasingly, individuals.
Companies know that when they hire agencies, they are paying for overhead and access to talent. An often frustrating aspect of this relationship for companies is that the senior talent they want is often quickly dropped and a junior slotted in place instead…the agency version of bait and switch.
This bait and switch issue and increased ease of finding strong talent that is freelancing, is the major reason agencies are suffering.
Smaller agencies are thriving because they have project management talent, can work swiftly and bring in senior talent as needed or work well with a freelancer the client company tells them to work with.
Within companies, they are shifting their marketing team roles to more project management and analytics based. This enables a company to more easily manage freelancers and multiple small agencies. They deploy tools like Slack or Trello and can leverage Office365 and GoogleDocs and now Dropbox’s Paper applications.
Most marketing and creative is about information creation and management. This largely negates the management architecture of large agencies.
Companies know how to find and leverage services like 99Designs, The Well, and many others. Talent can be anywhere. Agencies can’t control that anymore.

Busting Millennial Myths Around Social Media and Technology

Busting Millennial Myths Around Social Media and Technology
Note: This post also appears on my LinkedIn blog and The Well.
As a digital anthropologist, I research, mostly for marketing departments and agencies, how people behave within social media and with technology as a whole. Millennials are a hot topic. Many a brand spends inordinate amounts of time and money to target the Millennial. They think they’ve got them profiled, sorted and thoroughly understood and that they’re some kind of magical money mine. They aren’t. They’re also not really a demographic and can’t truly be marketed too as one.
How Millennials Are Using Social Media
In the past year I’ve completed over 15 different analysis of Millennials behaviours and activities in social media in the U.S., Canada and UK for CPG and financial services companies. So what did we learn?
  • They aren’t a “lump demographic”, they’re within an age bracket, but there are no specifics, just generalizations.
  • Over 87% of Millennials prefer social media apps that are less public; they dislike the marketing messages in more open apps like Facebook.
  • 59% of Millennials in our research indicate they mistrust over 90% of the news they get in their social media feeds.
  • When it comes to personal information, contrary to general assumptions, over 62% don’t like to share personal details outside a closed group.
  • Those aged 18-24 are more likely to be skeptical of unknown people attempting to be friends in apps like Snap or WhatsApp. But see Facebook and Instagram as more “open” platforms while sharing less information publicly.
  • Over 76% of those aged 18-35 say a primary part of choosing a new app is that it has the ability to have private messaging.
  • Over 56% of those aged 18-35 dislike the term “Millennial” and find it a derogatory term.
This is based on an overall sample size of n=75,000 individuals conversing about how they use social media in their channels in the US, UK and Canada between January and November 2016.
Millennials and Technology
While the media images of “Millennials” have them almost always face-in-smartphone, the relationship those in their 20’s and early 30’s have with technology is quite different from what one might expect.
  • Home automation tools like Amazon Echo or Nest aren’t really that interesting to them; largely because they don’t own houses and are highly urbanized and very much into “buying local” and a more social approach to shopping that is physical. Those aged 18-35 rarely have a house and don’t have as much disposable income either.
  • They’re 60% more likely to buy small commodity items on a mobile than via a laptop/dekstop or some home automation device.
  • Only about 12% of Millennials have an interest in wearables and tracking their fitness.
  • Just over 43% of those aged 18-35 say they don’t look for apps outside those pre-installed on the device they buy (i.e. the Mail app on iPhone or Outlook on Windows.)
  • 48% say they are looking to reduce the amount of technology in their daily lives.
We have a lot more data and insights than what we’re delivering here. But we think these are some pretty big insights. Millennials don’t like being called that and really, marketing to a specific demographic is a bit like trying to do a  direct mail campaign to unicorns.
What marketing departments tend to miss is that those aged 18-35 don’t have a lot of disposable income and in fact, have less disposable income today than their parents did at the same age.
Across many global consumer brands, we see this odd desire to leap on the bandwagon to market to “Millennials”, which is fine if you have a lower price point product. But price elasticity in the 18-35 group is much tighter. There are also significant economic disparities between those aged 18-25 (who have minimal disposable income and are in an entirely different life phase) and those 25-35, even those 30+ are very different in life stage.
Stay tuned for some further insights into how those aged 18-35 choose products and view brand loyalties.

How Blockchain Could Improve Aid Delivery

How Blockchain Could Improve Aid Delivery
One of the biggest concerns citizens and governments providing aid to developing nations, natural disaster hit zones and crisis situations have is the legitimate use of the donated funds. Corruption is a huge problem in international aid with some estimates arguing that less than 25% of every dollar donated actually gets used for its intended purpose.
Blockchain Could Deter Corruption
With blockchain, it becomes almost impossible to cheat the system. For a short explanatory video on how blockchain works go here. Essentially it is a lager system that is agreed upon by all in the blockchain network, any attempt to vary that system and the rest of the network rejects it. This means a transparent system, easily audited and all nodes no where the “money” or whatever value is assigned, goes. No closed bookkeeping in the blockchain.
So if an aid relief organisation such as UNOCHA or Medicins Sans Frontiers or Oxfam, or a government, donates funds (or even equipment or food) it can be tracked to the end use. With smartphones and scanners at the end point of where aid is delivered, corruption can be significantly mitigated.
The Economics & Politics of Blockchain Implementation
Implementing a blockchain system on this scale is not cheap, but conceding the cost of misplaced funds, food and equipment, it would likely be a lower overall cost. Initial costs would be set up of the system, ongoing management would be cost-effective. Aid value delivered could be increased to over 90%.
Political will is another story. Implementing it into developing nations and across their government and aid organisation infrastructure requires consent of the government receiving the aid. That’s a sticky wicket. In countries where democracy is non-existent or fragile at best, leaders can eject aid organisations at any time and sometimes do. How receptive would they be to a blockchain solution that avoids corruption and croneyism? It’s not an easy solution.
It will take time and if a fast, highly mobile and flexible/agile blockchain system can be developed, it could be deployed quickly in some situations. Some governments will welcome such a solution, others will prefer not to. But those governments that deny blockchain solutions for aid delivery will also look very suspect; not that they care.
Blockchain Takes Courage
A number of banks are researching the use of blockchain as are some financial services companies and corporations. But implementing blockchain solutions can be a challenge for many organisations. It is still seen largely as a novel and new technology by mainstream organisations. It takes courage to implement a blockchain solution.
Aid organisations aren’t known to be on the early adopter side of new technologies and blockchain is still in the early stages of the adoption curve.
What do you think? Is there a possibility for blockchain to help global aid delivery?

Why Facebook is The Worst at News Content – Napalm Girl

Why Facebook is The Worst at News Content – Napalm Girl
What the whole latest fiasco of Facebooks Napalm Girl and “news” feed and image policies really says, in the brutal truth.
  1. If Facebook is relying on AI to evaluate images, it shows just how bad Artificial Intelligence really is right now. All the AI Engine would have seen was a naked image…that’s what it was programmed to block, so it took the post down. This shows that AI is so very, very far from being good…since the AI system did not understand a) history b) relevance c) context d) emotion. Not much “intelligence” at work there. Facebook needs to train it’s AI algorithms better (and the folks writing those algorithms.)
  2. If Humans made the decision it shows that whomever is doing the hiring for editors at Facebook likely has a poor education and that the people they’re employing are even more poorly educated. Because they did not understand the historical significance of the photo. In addition, they were too daft to do any background research, like a simple drag and drop Google image search.
Either way, this shows that Facebook really does not understand a) news analysis b) journalism c) research d) current affairs c) international affairs. What it really says is “we hire twenty-somethings who like Starbucks and hashtags and are terribly educated” or “we’re using people as guinea pigs to train our aI algorithms that are really not quite ready for the real-world.”
This also says to me that Facebook really wasn’t playing preferences for democrat over republican content earlier this year. It says their AI algorithms are bad, or terrible, or that they hire inexperienced, uneducated people to manage their news content curation. If they do have a “senior editor” then that person likely is “senior” because they live in Silicon Valley and wrote over 100 blog posts about the nice weather in California and have an arts degree. Or the geeks overrule the experienced news editor.
At the end of the day, Facebook just isn’t a news content or publishing company. If they want to be, then they should take the time and effort and spend some money to bring on real, experienced (as in real, true journalists with deep international and domestic news experience) media people. Whether it be humans or machines running the “news” feed at Facebook, this latest incident proves they really do not have a clue about news content. If they have a highly experienced news person in a management role then that person is probably not getting the support they need and has a bunch of computer scientists and machine learning wonks overruling him/her/they…so how is that working out for you Facebook? Right.
So Facebook should either better support that news editor or they should hire a real news editor, let them build a proper team and the AI nerds should listen to that person. Then Facebook should make a point of promoting that awesomely skilled new journalism team to the world, showing their credentials beyond ordering complicated Starbucks lattes and having written ten blog posts and having an arts degree. Step up or step out Facebook. Right now, Facebook is entirely too opaque about the whole thing…that in itself says a lot. Your move Facebook.

We Need The Feminine in Artificial Intelligence

We Need The Feminine in Artificial Intelligence

Only recently have we seen the conversation around the need for more women being involved in the development of Artificial Intelligence. I argue that it is absolutely vital to have input and full engagement with women at the most senior level of development in Artificial Intelligence. No, this isn’t a feminist argument, it is a logical deduction and scientifically sound.

Since the early days of conceiving of AI in the 1940’s, development has been driven by men. All of our dystopian and quasi-Utopian Hollywood iterations of what AI might be like are driven by the male perspective. While the voices used in the voice-enabled aspects of AI are dominantly female, the underlying development isn’t. I think this is a problem for the future of A.I.

Why We Need The Feminine in Artificial Intelligence
I explore, quickly and briefly, the primary reasons we need the feminine as a key element in AI development;

Humanity: As humans we are male and female (setting aside the gender fluidity issue for a moment.) Ideally, AI would be gender neutral, but that is unlikely to be possible. To achieve greater balance and include “nurture” and the feminine perspective of what it means to be “intelligent” then the feminine needs to be involved. It’s the only way for AI as it becomes more advanced, to be closer to humans.

The Economics: For AI to be successful, it will have to prove its economic value. If AI is not profitable, companies will stop development. For AI to be more economically successful, it must include the feminine perspective. A singular perspective will lead to failure in business opportunities and profitability.

Qualia: Though we may never achieve this with AI, it is still an important consideration in the development of AI systems. The feminine experience of certain qualia is likely different from the male. So we see where the feminine is needed here as well.

More Than Data: Using data alone will not inform truly progressive AI systems. Experiences and perspectives are needed to train AI’s. The feminine needs to be considered here as well, otherwise we are again creating a system with only one perspective. That is not intelligent.

There is Some Progress
This year a small group of women will meet who 11 years ago formed a group called WiML and they’ll be holding their 11th annual workshop this year, coinciding with the NIPS conference. But so far representation for women in AI remains largely in academic circles. Industry needs to step up. Yet as the tech industry tries to and the likes of Bill Gates try to push this issue, just attracting women into computer science is a challenge itself. There are also some very intelligent women who are participating in the larger philosophical questions around AI’s development. But not enough.

In Summary
This is a short blog post attempting to posit a position in very short thoughts. While there is some progress, it is estimated that in the AI field, less than 14% of those engaged are women. This number needs to be much higher, but the tech industry is struggling just to attract women in the first place. AI may or may not become truly cognizant and aware and “beyond human”, but in case it does, we need to consider a lot of factors beyond just a “kill switch” and it may very well be that adding the feminine to AI at these early days may be what stops rogue AI’s from those dystopian futures that do so well at the box office.

Note: I consult at the senior executive level regarding Artificial Intelligence. My focus is on governance and risk issues, risk mitigation strategies, corporate strategy alignment and business case development.

What do you think?