Yesterday I wrote a lengthy ramble on what we need to change with the internet today. Much was about how we and others use our personal data to mislead us (propaganda, fake news, alternative facts, behavioral analysis and psychometrics), but then I realized that I have actually written about this before, more than a year ago. In the below post “Transparency is the future”.
At that time none of what I wrote yesterday was public knowledge (that does not mean it did not happen), but AI, cognitive and psychographic profiling (e.g. Personality Insight with Big 5 / OCEAN, sentiment etc) was starting to be seen as important. I then realized a big challenge in all the data that is available and what actually could happen if we did not think things through. The key takeaways are:
Transparency have long be pushed down in the value-chain in favor of capitalizing on our private data, companies like Facebook, Google and others that have as a business-model to re-sell your data to companies that want to reach you with their message. Their business model is to sell your data, you are the product.
This will not compute in the era of AI and cognitive. We will, in my opinion, be much more restricted in what we share about ourselves if we do not feel safe and secure.
We and everyone else in the segment need to be fully transparent about the following:
How we store the data
How we keep it safe
That personal information is not shared and are truly personal
Provide info on the reasoning on the answers provided
Provide evidence that backs the answers that are provided
Only then can a artificial intelligence and cognitive technologies really become as successful as they are expected to become.
Well, we did not think things through and now we have, misleading information and propaganda all over the place and our behavior online is constantly being profiled and then targeted…….but what is really sad is that we could have prevented it.
It is all our fault, we cannot blame anyone else!
… unfortunately, we did not decide to guard our personal data or to be transparent in what way the data is used!
It is all our fault, we cannot blame anyone else!
Photo: Kallbadhuset in Helsingborg (no I do not dip voluntarily in Swedish waters, too bloody cold), taken during a dog walk an early morning in March 2017.
With political powers, dictators raising, cyber-threats intensively increasing and nation sponsored propaganda machines and troll agencies always present, we need to improve several things with the Internet of today. We need to take back control over our personal data, stop misleading information and become more transparent.
It has been quite here on the blog for a while, there are several reasons for that, but one of the more important ones are that I have been struggling with what is happening with the Internet of today. The last several years, disinformation, propaganda and fake news / alternative facts have almost entirely taken over the flow of news and important information about the society we live in and the world in general. When you start to investigate this you realize how perfected the system is and how much impact it really has.
This is not about Brexit, Cambridge Analytica, Trump, Bannon, Sverigedemokraterna or Robert Mercer. They are just using the tools available at hand to reach their goals. It is how several components work together.
The System I am referring to here is much more than Trump and Russian (or US for that matter) propaganda and / or nation-wide hacking, it is more about how we consume information today and how the companies providing this information does not have any incitement to prevent this from happening, the opposite actually. The more times we click on these links the more money they make.
Facebook and Google make a living of presenting information they think we want, now that organizations, countries, brands, politicians and individuals have found ways to use this information to manipulate us, it is has become a flawed system.
Revenue makes this system tick
I will use Facebook as an example, but this is not isolated to Facebook in any way.
Facebook is an ad-tech company. Their business model is to provide tools for advertisers to put the advertising company’s information on the screens of a very targetted audience and charge money for that service.
User targeting. The success of Facebooks ad-tech is in two major parts. The first is the amount of personal data we have given to Facebook (does not matter how much you change your settings, they still will get to you) and the second thing is how they utilize that information to let advertisers target you based on interest, location, gender, age, civil status, your friends, what is in your photos, where your photos are taken, where you have been etc etc. An additional scary thing is that most have also given Facebook access to other personal data like browsing history etc, that is why you see travel ads on Facebook an hour after you search for your next vacation destination with Safari or Chrome, that info is added to your info used for targeting.
News targeting. This is what most people do not think about. We hear about filter bubbles and most of us have realized that the news we get is in some way personalized to fit my interest and behavior. Facebook wants us to interact as much as possible with content on the platform, so they push news to our feed that they think we want instead of showing a constant feed of lunch-pictures, which we do not interact with as much as we do with news-posts. That is why you see news articles more often even though just one friend has liked that post and 10 of your friends have more recently posted a picture of their newly ingested lunch.
Conversion. The conversion rate that Facebook provide in most of their metrics is actually astonishing and the reason for that is naturally no 2 above, but to further increase this conversion they are mixing no 2 and no 3 to elevate their revenue.
Since advertiser can impact what information that is put in front of the users, a bad spiral has been created (and that is why Mark Zuckerberg constantly denies that Facebook had any impact on the US election).
The scary part – Conversion
When it gets really scary is when this data is used to profile us and put information in front of us that we actually do not agree with, but when in the way it is presented we lean towards agreeing and by this changing our mindset. This can happen since behavioural analysis is done on us and information is then presented in a way to shift our mindset. This method is perfected and originally from Cambridge University (no affiliation to Cambridge Analytica though).
It is today easy to plant disinformation, propaganda or simply false news. This also creates nationalism and authoritarianism.
This is the system that Cambridge Analytica has been utilizing when they helped Trump get elected. They did social profiling on almost every US citizen that can vote and personalized their feed to shift their beliefs (and if it did not work re-iterated until it did, on individual basis) in favor of Donald Trump. Cambridge Analytica is a company rarely seen in public, but a brief background is that it is funded by billionaire Robert Mercer from one of the most successful hedge funds in the world, Renaissance Technologies (before becoming co-CEO of this hedge fund he worked at IBM Research on computational linguistics). Cambridge Analytica also has senior Trump advisor Stephen Bannon on the Board of Directors. The CEO is Alexander Nix and it has a UK version of it as well with the name of SCL Group. SCL Group did the same thing for Brexit. Below is a video of Alexander Nix presenting what they did for Ted Cruz in the primaries, before they joined forces with Trump, it still describes what they do pretty well.
The evil ecosystem
This ecosystem is several parts that feed off each other.
The Platforms (Facebook etc)
The Impactors (I think I just made up that word). The ones that want to impact us
The users, that would be you and I
The platforms want revenue, the impactors want to impact us, the users want news and relevant information.
As you see in the ecosystem section above the last part is flawed, since we are not getting information and news that is relevant, but information others (the impactors) want us to consume and be impacted from. This is a major flaw on the internet today. It impact us all in a very very bad way.
As most things in business, Facebook (again, they are just the easy to use example here) is revenue driven and they want us to see and potentially interact with as many ads as possible and that makes us vulnerable for misleading information that creates a world made of powerful people where we, the internet users are the enabler for this happening, this without our own intent and knowing.
We need to stop misleading information and take control over our personal data.
Why this rambling? I have for some time tried to get my mind around this and since I have been quite heavily into cognitive computing and AI the last years a lot of thoughts have crossed my mind. With my background it might have been logic to work more intensively with large corporations on how to target their audience in a better way with behavioral targeting (using cognitive solutions), but my conclusion is that it is not ethical to use social, behavioral profiling and cognitive computing to do this.
I have at least 10 posts on this topic in draft mode, but have not published since it only paints a negative picture of the current world and some are almost doomsday prophecies. Just imagine the post I have on “Cognitive Warfare”, interesting, but also a bit scary, will keep it in draft for a while longer.
We need to tackle the problem with personal data and misleading information first.
When the users are safe and the information is not misleading and transparent, first after that we can use cognitive to interact with users the right way as a business.
On oh so many services you do find the “Find My Friends” button that will help you get better content in service you use. You find those Facebook and Twitter logins everywhere that want access to your friends to bring you a better “experience”.
I find it mainly a flawed system. This due to the simple fact that the people I follow on Twitter and the friends I have on Facebook etc, might not be the ones I share interest with on the new service. I’ll just give two examples:
Spotify. Spotify launched it’s social discovery feature a long time ago and did it together with Facebook. The thing is that my Facebook friends do not listen to the music I want to discover, they are mainly my friends on Facebook for other reasons.
Pocket. Pocket is my favorite “read-later”-app. It just launched a follow friends feature, I tried it out the only result was that I found the same links that I had already seen on Twitter, from the same people that I follow on Twitter and now in Pocket as well.
There are many many more examples. There are occasions when this do work though. I love both the Newsle and Nuzzle services as examples. Newsle finds article and posts about my connections on LinkedIn, but also from my mail-contacts. Nuzzle simply digests the most intereseting posts from my network from a specific time based on shares etc within my friend / follower network, I do have Nuzzle deliver a daily digest every morning with the last 24 hours hot news in it.
Is there a solution to this flawed “Find my friends” thing?
Naturally there is. I actually took both the above examples since they have alternative ways or actually have changed there service since it launched.
The solution is to know the individual better , learn from the user and then from that, use the data at hand to discover top content for the user.
Spotify naturally uses its own technology as well as the acquired Echo Nest platform. it uses natural language to understand blogs, titles and meta data. Then there are many other machine learning things as well as Kafka to work with the data in real-time.
Compared to the the list of what my friends on Facebook listened to, this is huge progress and today a big portion of my listening comes from either curated or automated playlists on Spotify. My user behaiviour have completely changed since Spotify started to work dedicated with Discovery.
So I am not a fan of Pockets “What friends share” in the Recommended section, but I am a fan of the posts that Pocket itself recommend to me. It is based on my reading and then presents personal recommendations for me. It is a much simpler use case then the above Spotify one. Pocket also uses natural language processing and simply uses the IBM Watson AlchemyLanguage API on its content. They look at my content and let the Watson service digest entities as well as extract conceps from that content and then use those to find other popular content that matches that info. The wast majority of recommendations I get from Pocket in the Recommendation section (not the ones shared by friends I follow) are spot on and I would say that about 50% of my saves in Pocket has originated from that part of the feature.
Cognition is the set of all mental abilities and processes related to knowledge, attention, memory and working memory, judgment and evaluation, reasoning and “computation”, problem solving and decision making, comprehension and production of language, etc. Human cognition is conscious and unconscious, concrete or abstract, as well as intuitive (like knowledge of a language) and conceptual (like a model of a language). Cognitive processes use existing knowledge and generate new knowledge.
In conclusion, we are finally entering the era of thinking and awareness (which is the original latin meaning of the word), even for computers.
A cognitive process can be divided into 4 steps
Decide (action, present confidence and probability etc)
Research to dive deeper into cognitive
Cognitive is not only a marketing term for selling Watson, it is actually something researchers have worked on for a very long time, this since quite a few IBM researchers work on how the brain works. In the context of this post it naturally gets really intersting when computers and software starts to adopt cognitive functionality. .
According to Arvin Krishna, Senior Vice President of IBM Research, a cognitive technology needs to:
Learn at scale
Reason with purpose
Interact with humans naturally
Has an objective (this bullet was actually not said by Arvin but by Dr John Kelly in a talk at Cognitive Collaquium NY recently)
This is crisp and easy to understand (even though understanding how the hell they got the software to work like that is another thing).
IBM have always been very generous about posting their research online. It can be tricky to find, but it is out there. In terms of cognitive it has its own section on the IBM Research site. So either start on the IBM Research Cognitive Computing section or at the IBM Research homepage. There is droves of info to read up on and hours of video to watch.
Cognitive analytics A set of technologies and processes that analyze data for the purposes of learning, contextualization, and making recommendations.
Cognitive business An IBM strategy that builds on digital business and digital intelligence with systems that can understand, reason, and learn to leverage data to create deeper engagement and personalization, enhanced expertise, and cognitive products, services, operations, and processes.
Cognitive computing A category of technologies that uses natural language processing and machine learning to enable people and machines to interact more naturally to extend and magnify human expertise and cognition.
Cognitive environment An infrastructure that uses specialized software agents and devices that act as one shared integrated resource, enabling fast and efficient human-computer collaboration.
Cognitive system A category of technologies that uses natural language processing and machine learning to enable people and machines to interact more naturally to extend and magnify human expertise and cognition. Watson is an example of a cognitive system.
This is also the second time I borrow a top picture from IBM, this one from their campaign-site “Outthink” that is out there to enlighten us about Cognitive. It is a typical campaign including a darn cool commercial with Bob Dylan talking to Watson..
We hear about the IBM Watson everywhere. It began with Watson winning Jeopardy in 2011, then the message and the technology behind has been pollished to now being commercially viable both as strategy, technology, marketing, and now also as a busienss.
The IBM story that led to IBM Watson
Let’s start with a little flashback. Generally IBM is perceived as best (only) suitable for large and bulky corporations. It is in fact a very true reality-based picture. IBM, however, has driven innovation in a way that few other companies have. Fluffy concepts that few have understood have after a few years become a reality, and IBM has rarely received any credit for it, despite the fact that they pushed much of the early innovation and research in the field. Terms such as e-business and on-demand were both coined by IBM. IBM has had its fingers in most big shifts in technology over the years.
It began with the tabulating machines and punched cards. The tabulating machines and punched cards are kind of the first computer ever made.
Programmable systems. IBM with their computers and their databases, programming languages and operating systems have helped to create this era, it is quite natural. It was IBM who introduced the magnetic hard drive, DRAM, etc., etc., but also the relational database in 1970 and that they actively contributed to several programming languages.
There are of course several other companies that have affected the industry over the years, but I argue that no other company has done it the same way as IBM. Today, IBM is the world’s largest IT company (usually # 1, but it shifts between IBM and HP). IBM has about 400,000 employees and are in the leading position in most areas they operate in.
The toughest time for IBM was during the 1990s, when IBM was very heavy and was stuck in history. Some of the same challenges we see that IBM have today.
The share price today is the lowest in five years, analysts thrashes quarterly reports and the future success is constantly questioned. Is IBM a dead horse?
I’m not an analyst and do not want to be either, but in my world, the stock market is a rigged game, it is financial models more than if companies are doing well or not. IBM is the only listed share I own and I doubled up my little possession as late as last week. Why? Well, I believe in IBM and I think Watson is a big part of IBM’s future success.
IBM suffer from being big (big blue) and have been around for over 100 years, although the IBM name came first in 1924. It all started when Charles Flint merged together four companies, and in 1911 Thomas J. Watson, Sr. became General Manager of CTR, IBM name at that time.
This allows IBM to drag on a lot of legacy. It’s almost a little scary when you read in the business press and comments from analysts that IBM is a hardware company. It was a very long time ago IBM had hardware that some kind of core business and today it is just over 10% of revenue coming from hardware and margins I guess are at a minimum, which IBM realized in time and sold off most to Lenovo. The cash-cow mainframe is kept and doing very well as far as I have understood.
Why are the analysts wrong?
IBM is today in terms of sales about 50% Consulting, 35% Software, the rest is small numbers. What should worry the analysts, if one should find something to complain about, is that the kind of consultants that IBM is offering is moving toward decreasing demand in a world where companies buy more and more SaaS solutions, those solutions are not in need of lots of consultants in nice suits for extremely expensive hourly rates. Sure, I generalize, but the fact is that the demand decreases, and right from the start, consulting is no high-margin product, so personally, I see a transformation like Lou Gerstner did, when he shifted from hardware to consulting 15 years ago. For those interested, his book “Who says elephants can’t dance” is a really good read about that transformation.
Just to be clear, both GBS and GTS have a lot of geat offerings which are still very valuable and affordable for customers including outsourcing. It is the actual hourly consulting I refer to.
Software, on the other hand, is about 80% margin (GP Margin 3Q15 was up 86.4%). Software is the new black, but there are clouds in the sky, the license-model that IBM offers are from the stonesge. IBM’s licensing model is based on selling large volumes at discounted price (this is strucutred in a system, called Passport Advantage) and with volumes the customer can grow with, that they do not need now, but maybe in the future, aka the shelfware. In Sweden (that I now have retured to after 3 years in southeast asia), I would guess that these transactions corresponds to approximately 80% of the revenue and only comes from about 20 customers (maybe should have added a “Hello ELA” in the title). It is not the future!
IBM have had the SMB sector as a major focus for about 10 years now, unfortunately, it is nice words than operational reality, IBM goes where the money is and in their case that is the big corporations. Even here I think that Watson will be able to bring a change, both in the business model as well as he products.
Since I wrote above from the top of my head, I felt that it might be worth checking facts about the numbers, so I checked the IBM 3Q15 earnings presentation from October 19, and my numbers correspond well with the 3Q15 earnings numbers. It again illustrates my skepticism towards the stock market. The overall profit margin for IBM is impressive 50%, it’s a pretty nice margin. If you move to higher margin products which require less human effort and away from products that generate the most revenue (hour-based-consultants and hardware) it will increase margin but sales go down, why do not the analysts see this, but what do I know, I’m just thinking logical and that is the opposite of the stock market.
But are the analysts right after all? Oh no, they’re wrong, I say optimistic, but there are some challenges? A look at their product portfolio and ignore the different IBM divisions, one sees quickly that IBM must manage to turn its business model and its products to the cloud and valuebased subscription models. Sounds buzzwordy, but it is crass. With competitors in the cloud as Amazon, but also Google, Microsoft and Salesforce et al. Personally, I think that this particular infrastructural change in terms of the cloud, will be the toughest challenge of them all for IBM.
With the above in the rearview mirror, I’m very happy to see the intense focus IBM management has on Watson. Every time IBM CEO and Chairman Ginni Rometty speaks she always mention Watson, and when you see Ginny it really feels that she is passionate and sincerely believe in Watson and its development.
Watson is not a temporary marketing fix, it is IBM’s future, or at least a very important component in IBM’s future.
My personal opinion is that she invests her name, her success as well as IBMs in the success of Watson. I believe Watson to be the single most important part of IBM to be successful long-term.
Why is Watson so important to IBM?
Why this exposition on the share price, turnover, success with cloud etc? What does that have to do with IBM Watson? Well, Watson represents a new era in technology, so if we link together IBM’s history with this new era, the cognitive era, this picture is very simple and descriptive about the different periods we had and the one we are now entering.
But it is not just by what Watson can do in cognitive analysis, but also the business model. Watson lives in the cloud, has an extremely flexible and adaptable business model and that Watson is not only aimed at large companies, but also for startups and others. For me Watson represents IBM’s future at all levels.
Watson’s success will be based on Watson’s part in other companies products and that requires a good genetic price / licensing model, and that services are available in the cloud. There we have a future for IBM that I believe in.
This post was intended as a “Hello World”-post in terms of Watson, but now when I have written it, it should rather act as a philosophical baseline of my attitude towards IBM and its future. I have a few upcoming posts on the topic and those will guaranteed be more related to Watson. It all sticks together though since Watson is IBM’s future. IBM Watson Group with about 2,000 employees has now existed for almost 2 years, the recently founded a consulting unit of GBS, IBM Cognitive Business Solutions, with about 2000 consultants who solely will work with cognitive and further 25,000 IBMers are to be trained in the field this autumn, over time all IBMers will most probably get their fair share of Watson.