Thursday, October 22, 2020

Too many Covid Experts - Whom should we trust?

 The answer seems to be: nobody!

With Covid raging on, we are all living under fear and stress. Countries such as China, New Zealand, or Japan that succeeded in keeping it contained, are hunkering down behind Chinese Walls, while the rest of the World is desperately searching for miracle cures. Our daily life is heavily influenced by coping and surviving under Covid19. Self-proclaimed Covid19 experts are inundating us with a never-ending stream of news and insights. These experts come from two opposing sides: mainstream science and government experts on the one side, and conspiracy theorists on the other side.  Figuring out whom to believe, with so many “experts” contradicting each other can become a real headache. Politicians, regulators, and scientists in different countries, and even within the same country, are fundamentally disagreeing and fighting with each other about the best strategies for coping with the disease. 

To shed some light, we did a coolhunting using Galaxyscope, creating two digital tribes, “Covid-Experts”, and “Alt-health” (spiritual healing believers, Covid deniers, and fringe conspiracy theorists). The picture below shows the two networks. 


While the Covid-Experts form a solid cluster, Alt-health has some separate clusters, connected by a few gatekeepers. However, the most shocking insight is that DrTedros the director of the WHO, the World Health Organization, who should take the lead in the fight against Covid19, appears in both networks – at the periphery. This means that mainstream science and spiritual healers are both desperately looking for leadership – and not finding it.
In the Covid Experts community, people like Eric Topol, Director of the Scripps translational institute, Helen Branswell, a Canadian global health reporter, and Marion Koopmans, a Dutch virologist, occupy the central positions. In the Alt-Health cluster, activists like FLAutismMom, Brett Weinstein and Spiro Skouras are in the center.
The picture below shows the two Twitter word clouds of the two tribes, with the central term “Covid19” removed, because it would overly dominate the word cloud.  While the experts talk about approaches for fighting Covid such as contact tracing and NIH, the alt health members talked about the Fox TV show “the masked singer” which became highly popular as a way of coping with Covid-related stress, but almost went under in the summer when its host Nick Cannon made anti-Semitic remarks.

The picture below shows the main words used by experts and alt-health members as word networks, where a connecting line between two words means that the two words appear in the same tweet. Experts recommend wearing a mask as the key Covid-prevention measure. Alt-health members lead a far more wide-ranging discussion with different word clusters, the Bill Gates conspiracy is in the center, and attacking Dr Fauci, the main US government Covid expert. Trump, Maga, and the “proud boys”, appear in the periphery.


Looking at the tweeting activity on the charts below, the experts are far more active tweeting about Covid than are the alt-health members. Alt-health members tweet more about non-Covid topics related to conservative politics, such as bikingthebattleground of conservative commentator Cheryl Chumley. 

Looking at the emotions associated with Covid, there is no noticeable difference between Covid experts and alt health. Fear and sadness dominate for both experts and alt health members in the emotion charts below. Happiness is low for experts and even lower for alt health members, and fear is high for experts and alt health members alike. 

In the comparison charts of the two tribes below, we see that while experts tweet much more, the two tribes have little overlap in membership, although some Covid experts are positioned right among the alt health tribe members by our positioning algorithm. The dark blue dots at the bottom of the scatter plot below at right stand for Michael Eisen, a biologist at UC Berkeley and Seema Yasmin, an epidemiologist, who use language in their tweets more typical of alt health.

Looking at the membership in predefined virtual tribes below, we find that the experts are more treehuggers, while the alt health members are also fatherlanders (ultrapatriots), while there are none among the experts. Some alt health members not surprisingly have been categorized as spiritualists.

Personality-wise, the experts use a lot of political and journalist language, while the alt health members are more categorized as risk-takers and stock traders.

Covid experts are more capitalist than alt health members, which is not surprising as they work for the US government, pharmaceuticals, and academia, while some alt health members are seen as socialist in the way they speak.

In conclusion, we see that fear and sadness dominate the Covid19 discourse, with experts recommending handwashing and mask wearing, while the alt health members see conspiracies everywhere, and recommend miracle cures against Covid such as Melatonin and Vitamin C instead of having to wear masks. Both groups place strong emphasis on maintaining mental health among all the chaos. The main insight is that there are no accepted leaders, and very little trust in science. Everybody seems to hunker down and hope for better times – which undoubtedly will come, as they always have, we just don’t know when!

Many thanks again to Karin Frick from GDI for doing the coolhunting.

Thursday, October 15, 2020

Using People Analytics for Analyzing “People Analytics” Thought Leaders

Increasing workplace satisfaction and recognizing employee burnout before it happens are core issues in today’s Covid-19 home office workplace. People analytics has become a key technology to support these goals, thanks to recent progress in AI, machine learning, and other computer technologies. Using our own people analytics tools we analyzed the virtual tribe of key people working in the area of people analytics. If anything, this analysis shows that people analytics has become even more central in the Covid-world. 

Using our Galaxyscope Tribefinder tool we created a virtual tribe of 57 Thought Leaders in the space of HR analytics, people analytics, and corporate culture. The picture below shows the social network of the people analytics thought leaders based on their Twitter network. Thought leaders are connected through who is retweeting and mentioning whom. It also categorizes each thought leader into an “alternative reality” tribe, identifying if somebody is a fatherlander, nerd, treehugger, or spiritualist (. As we can see in the picture below, the large majority, 38 out of the 57, are nerds, based on the language they use in their tweets. This is not surprising as they tweet about technology.


Among the 57 people who are popular tweeting about people and workplace analytics, we find that the most central people analytics thought leaders are David Green, Wharton professor Adam Grant, ABN Amro HR analytics manager Patrick Coolen,  Andy Spence, and Josh Bersin, to name just a few. A connecting line in the network means that two people are retweeting or forwarding or mentioning each other.

The word cloud below shows the most frequent terms in their tweets of the last two months. “HR”, “HCM” (Human Capital Management), “people analytics”, and “future of work” are most popular.




The next picture drills down into the word cloud by showing the links between two terms, if they are mentioned in the same tweet. The word network nicely illustrates how Covid19 has created a second separate word cluster about the leadership of remote teams, which is a key issue in the home office workplace mandated by the Covid19 lockdown.




When looking at the popularity of hashtags over time, shown below, we find that the importance of “culture” is growing, together with “HR”, “future of work”, and “HCM”.


The last chart shows the characteristics of the tribe. It very well reflects the personalities of the people analytics thought leaders as a group of nerds. They are mostly fitness nuts, like to travel, are ingrained in capitalism (not surprising as people analytics is popular with large multinational corporations), but they are also risk takers, writing about early trends ahead of the mainstream.



For more information about how tribefinder and galaxyscope works, see my earlier blogpost and the papers on www.ickn.org, for instance this one. Many thanks to Karin Frick from GDI for doing the coolhunting.
  

Wednesday, September 02, 2020

How to Find Interesting Research Problems?

 This morning I was asked this great question, which I am struggling to answer, as this is a problem with many dimensions. Anyway, here is my try to give a partial answer.

The first answer is “never ending curiosity”. I wake up in the morning just wondering why things are how they are. Why are trees growing towards light, and not towards darkness? Why do we need light to live? Photosynthesis describes a pattern, but does not give the fundamental answer. There is always so many more questions than answers, and every answer brings more questions.

The second answer is “I want to find answers to questions I am passionate about”. So, you need to find your passion. How to find your passion? One answer is looking at other people. Whom do you admire most? What are people whom you find cool doing? For instance, Elon Musk said that he was most inspired by Nikola Tesla, as a consequence he started exploring electricity, ultimately building electric cars and naming his company “Tesla”. The area of research you choose also has to do with your personality. I noticed that an economist, a marketing researcher, a computer scientist, a psychologist, and a zoologist have very different personalities. The personality differences are even larger between soldiers, physicians, professors, entrepreneurs, and managers.

If the most exciting thing for you is being a professor at a top university, then choose your research questions opportunistically. Try to find answers to “hot” research problems in your area. For instance, a hot area right now is fake news detection and prevention, using AI. 

Personally, I am much more interested in “cool” things than in “hot” things. The difference between “hot” and “cool” is that “hot problems” are in everybody’s awareness, which means it will be much easier to get your paper accepted, if you offer an incrementally new solution extending an existing solution. “Cool” problems are ahead of their time. There is a small group of other “crazy researchers” working on them, but it will be much harder to get papers accepted about “cool” topics. Papers addressing “hot problems” will offer incremental innovation, papers addressing “cool problems” will offer radical breakthrough innovation. The problem is that at first people will say it’s crazy, does not work, or is not a problem worth investigating, until it is suddenly “obvious”. I have seen that many times. For instance, in 2003 or 2004 I was in a meeting with VCs who were discussing investing in face recognition startups. They had invited a famous professor from a top university, an expert in image recognition, who told them that computers would never be able to recognize faces as accurately as people could. Well, fast forward a few years, and computers have become much better than humans in this task – isn’t that obvious! 

The pebble rolls down the hill, inspiring the concept of the wheel. Innovation consist of applying existing solutions from other fields to the problem at hand. How to find new ideas? Talk to as many smart people from as many different backgrounds as possible. This is why I like to teach at universities around the world, students in Finland, Chile, the US, China, Italy, Germany, etc. have very different perspectives of the same problem. In my research I am straddling mathematics, computer science, management science, sociology, psychology, philosophy, and biology which I find immensely enriching. As is bringing computers to children in the developing world, working on reducing infant mortality in the US, and trying to understand how plants, horses and dogs communicate. In the end this will even inspire a better understanding of how fake news spread. 


Sunday, August 30, 2020

AI-Enhanced Interspecies Communication

 While we humans have a hard time communicating among ourselves, we find it even more difficult to communicate with individuals of other species. The main interaction between species consists of eating members of other species, or being eaten. We humans consume huge amounts of pig, cow, and chicken meat, drink milk, eat eggs, bread, rice, corn and other veggies. However, when we are not eating plants and animals, we also like to talk to them. Some people even talk to their houseplants. And owners of dogs, cats, and horses of course talk to their pets all the time, maybe even considering them their soulmates. But do the plants and animals really understand what we are trying to say to them? And even more important, can we understand what the dog, horse, cat, or a mimosa or basil plant is trying to say to us?

Before we can start talking to others, we need to listen to what they have to say, and try to make sense of their output. Computers and artificial intelligence have made huge progress over the last twenty years helping us both to listen and to talk to each other. Today’s New York Times gives a great overview of using computers to read the brain which explains how computers are capable of knowing what we are thinking by tracking the activation of combinations of brain cells indicative of certain words. In the currently best implementations computers can read up to 250 different words at 90 percent accuracy by looking at our neurons. In genetically engineered mice, computers have also spoken to the mice, telling them when to drink water by turning on their neurons, “playing them like a piano”. Google Translate and Deepl have become language geniuses, dynamically translating from English to Chinese, with me talking to my phone in English in the hotel in Beijing, and the phone repeating my sentence to the hotel receptionist in Chinese.

What if we could do the same when talking to a horse, a dog, or a mimosa, with the computer telling the animal or plant what I would like to say, and then the animal or plant talking back to me.

Our research group has been studying how humans communicate online and face to face for the last two decades. In our work we have been building many tools for happier, more creative, and more productive collaboration among humans, leveraging the computer and AI to read the "honest signals" and emotions of what a human really wants to say beyond the literal meaning of words.

In the last few years we have been applying these algorithms and technologies to interspecies communication. We have used body sensing technology to better understand communication with horses, and facial emotion recognition to understand emotions of dogs and of horses  . 

In our most recent projects we are trying to listen to plants. We have put “brain sensors” on mimosas, as they talk back to the outside world quite visibly: when their leaves are touched, they fold them. We found that they seem to sense the electrostatic discharge of human bodies and the rhythm of body movement near them. When putting our sensors on other plants such as basil they show the same response. We also recorded the leaf movement of the “dancing plant” (Codarialcalyx Motorius) in response to human voice and music. Using automatic image recognition we found that its leaf movement was different in reaction to male and female voices talking nearby.

As this is a brand-new emerging area of research, we are at the very beginning. Imagine how wonderful it will be not just talking to your dog, cat, horse, or houseplant, but being talked back in understandable words, and engaging into a real dialog. 

Much more research is needed, if you are interested in collaborating in our research, we love to hear from you. 



Monday, June 15, 2020

Is Covid19 breeding humans for optimal dissemination?

Noah Harari in his book “Sapiens” makes the argument that humans make pigs, chicken, and cows the most successful species by breeding them in immense numbers and thus creating a huge gene pool. Humans are doctoring around in this gene pool, for instance breeding chickens to gain weight rapidly for more meat production. But as these poor chickens can not even walk anymore, this gene doctoring leads to evolutionary dead ends.

Coronavirus seems to be doing the same to humans, it is treating us like we are treating the pigs, breeding us to provide the best possible living environment for Covid19. Towards that goal, it kills the least attractive hosts, old people that will not live much longer to spread the virus, bald men, which have too much testosterone, and overweight people who frequently have preexisting conditions such as diabetes, high blood pressure, and lung disease.

It seems thus that Covid19 is weeding out the weak to improve the gene pool of the human race for further successful dissemination of the virus.

What does this say about Brazil and the US having huge infection and death numbers, and Germany having comparatively very low numbers?

From an evolutionary perspective, people with poor health such as obesity, high blood pressure and diabetes are also predominantly poor and the least successful economically, and that is why the Coronavirus is attacking them the most.

As empathic and compassionate human beings, what we have to do is to level the playing ground, and give poor people the necessary support to be resilient against failure, then their health will also increase and Coronavirus will not be the killer it is for them right now. It has been shown many times that economic success and good health are strongly correlated. In particular, the theory of ACEs demonstrates that having a supportive childhood is the greatest predictor of adult success. Therefore we should try to give all children the same high level of physical and emotional care right from when they are born. This is for instance the goal of the Healthy Start and Infant Mortality CoIIN projects in the US.

Tuesday, May 26, 2020

My Truth is NOT your Truth - Predicting Truth Based on Ethical Values

What is “truth”? 
Truth is highly subjective and depends on the individual context, likes, personality, and ethical values of a person. What is “truth” for you depends in whom you trust and in whom you believe. If you are a farmer and Republican voter of Donald Trump in the rust belt of the US, you believe China needs to be punished for ripping of the US and you trust the words of Donald Trump and are willing to sacrifice sales of your grain or pork to China for the greater good of your country. If you are a Chinese Communist party member, you believe the US is striving for World domination and is using whatever means available to subdue China, you trust the words of Xi Jinping and are willing to bring personal sacrifices for the greater good of the Chinese nation. The same split perception of what is “true” applies to many other issues, from the effectiveness or dangers of vaccines, miracle cures against the Coronavirus, to when the world was created (4.54 billion years ago according to science, less than 10,000 years ago according to creationists). As nobody can actually go back and check for themselves, we have to trust and believe either the scientists, or the spiritual leaders of the Christian fundamentalists. To back up their claims, both the creationist leaders and the scientists show us long chains of “facts”. “Facts” can be photos of fossilized dinosaur feet stepping over human feet, proving according to creationists that dinosaurs and humans lived at the same time, about 8000 years ago. “Facts” can also be the physical formulas that prove the existence of spacetime curvature predicted by Einstein. Both of these “facts” are impossible to verify for the person on the street, and so believers of a particular “fact” take them at face value thanks to their trust in the people coming up with these “facts”.

Ethics and Personality Define How to Interpret Somebody’s Claim to Truth
Depending on our motivation, we will apply a different definition of truth to solve a particular problem:

  • Donald Trump’s mantra is “truth is what gets me re-elected”. If he says that hydroxychloroquine will cure Corona virus infection, this is because he assumes that having a cheap and easily available "miracle cure" will increase his chances of re-election.
  • A sales guy’s mantra is “truth is what sells my product”. If a pharma vendor of hydroxychloroquine says that hydroxychloroquine will cure Corona virus infection this is because the sales guy wants to sell more of his pills.
  • If a doctor tells a patient that hydroxychloroquine will cure her Corona virus infection this might be because this is the only drug he has available to help the sick patient.

What Donald Trump, the sales guy, and the doctor do, is closely tied to their ethical values.
It has been shown that particular personality characteristics go along with certain ethical values  and professions. For instance, using the FFI personality test, agreeability and conscientiousness is prevalent in highly religious people, while academics show very different personality characteristics: A scientist will be open to new ideas, and willing to question existing beliefs and authority, but will be much less agreeable and less conscientious. A creationist will value tradition and authority over progressive ideas. As “birds of a feather flock together”, creationists will all believe similar facts, and they will show similar personality characteristics. In other words, they will flock together in “virtual tribes”.

4 Main Belief-Systems define alternative realities
In an earlier blogpost I introduced a belief system based on digital virtual tribes which has now been extended to four categories:

  • Fatherlanders – like to maintain the status quo, they do not like change, and they very much distinguish between “us” and “them”. Not surprisingly, fatherlanders show the same characteristics as highly religious people. Usually their value system combines “god and fatherland”. On a side note, over the course of human history more people have been killed for “god and fatherland” than for anything else. 
  • Nerds – are believers in science and progress, looking for solutions to their problems from new insights in science and technology. At the same time, many nerds are also shrewd investors converting their technological savvy into personal fortunes, just look at Bill Gates, Larry Page, Sergey Brin, Mark Zuckerberg, and Elon Musk.
  • Treehuggers – are environmentalists fighting for the protection of nature and the environment, against global warming, and against unfettered growth and robber capitalism.
  • Spiritualists – believe in supernatural forces, and in the power of their own mind to achieve a higher level of awareness, mindfulness, and happiness.

A challenge to a deeply ingrained belief will only be accepted from an established leader in the same belief-realm.  This means that if a Nobel-prize winning scientist would tell another person, another nerd with a similar personality and belief system that “the world has been created less than 10,000 years ago”, this would be far more credible to the nerd than if a fundamental Christian would say the same thing. On the other hand, fundamental Christians will accept the claim that the world was created less than 10,000 years ago as a fact, as it is supported by their spiritual leaders, while they don’t really trust science, speculating that many claims of science might be just a ploy to justify world domination.



As the picture above shows, the ethical values and personality characteristics define if somebody is a fatherlander, nerd, treehugger, or spiritualist. And if we know to what belief category (fatherlander, nerd, treehugger, or spiritualist) somebody belongs, we will know which rule somebody will apply to come to truth about how the world was created. A fatherlander, who usually is also highly religious, will take the book of Genesis in the bible at face value, and thus accept that the world has been created in seven days. A nerd will trust scientific cosmology and accept a creation date of the universe 4.53 billion years ago as truth. A treehugger might accept a creation date in the very distant past, but might show skepticism towards science and care more about how to keep nature intact, accepting global warming as truth. The spiritualist might similarly distrust cosmology, but will accept that some supernatural forces created the universe a long time ago, with these supernatural forces still ruling our behavior today.

The Belief System Predicts Individual Behavior
To predict how somebody will react, we thus will need the believe system of a person. If I know your personality and your ethical values, I will. know what is truth for you. The picture below shows how the four different tribal belief systems lead to four different truths about curing a Corona virus infection.


Knowing your virtual tribe we can predict what you will do if you fear you are down with the Coronavirus: if you will go buy Remdesivir, you are a nerd; if you will pray for healing, while maybe swallowing a few pills of Hydroxychloroquine, you are a fatherlander, if you will scour the web for herbal remedies – a Madagascan herb (artemisia annua) seems a current favorite – you are a fanatic treehugger, and if you muster the inner forces of your body through meditation to fight the virus you are a spiritualist.

Can we Discover your Ethical Values and Personality?
It would be great if we could put a magic bracelet on anybody, and then know what their truth is – this will also relieve their need to lie, as their magic bracelet will show their truth to the world anyway. Well, there is such a bracelet. Over the last four years, based on prior work using sociometric badges, we have developed the Happimeter, which measures the body signals of an individual such as heartrate, movement through accelerometer, and speech patterns (no content) from sound, and location changes through the GPS. It works with smartwatches (the current version runs with Android Wear and Apple Watch) and will predict your personality characteristics and ethical values.

As second way to discover ethical values, personality, and belief system of an individual is through the words that one uses. We have shown that word usage in e-mails predicts ethical values, personality characteristics, and tribal belief system.

This means that using AI and machine learning, the way how you talk, and the way how you move your body will predict whether you say the truth – that is, what you believe is the truth, and nothing but the truth!



Saturday, May 16, 2020

Is the next Einstein an AI System?

Technically, AI can now do many cognitive tasks much better than humans, molecular modeling is one of them, NLP and image recognition are among the most well-known. But will AI become truly creative? Or in other words, can AI predict unknown unknowns. We already know it is great in discovering and solving known unknowns?

Philosophically speaking the question is if the next Einstein will be an AI system. The answer is most likely yes. The saying is that humans are unbelievably complex. I have found that humans are extremely simple, always applying simple heuristics. The point is that we are not aware of these heuristics and come up with super complex models and explanations, leading at times to wild conspiracy theories. Mr.Trump is embodying this with his wild theories, ranging from vaccination theories to Obamagate, while his simple heuristic is “do what will most likely get me re-elected”, unimpeded by any ethical constraints and restrictions.

We just need to apply Occam’s razor “the simplest solution is most likely the right one”.
Gerd Gigerenzer  has written a series of books, the most well known is “Simple Heuristics That Make Us Smart”, where he explains that many complex problems can be much better solved with simple heuristics,
In my own research over the last twenty years I have been trying to uncover as many of these really simple heuristics as possible. The honest signals of collaboration are some of them, digital virtual tribes is another one.

Once we have found all of these simple heuristics, we will teach them to the AI system, which then can take over and be the next Einstein. But this is still a long process, as it is awfully hard to discover these simple heuristics.
In the meantime I think we still need to try to obtain a deep understanding of each problem, to identify the underlying simple heuristic that solves it. In the case of Einstein himself, he was using his famous thought experiments to come up with simple and intuitive solutions to fiendishly complex problem that other physicists had been unable to crack. For instance, Einstein was famously riding an elevator in the patent office of Berne, when eureka struck him and he saw the principle of relativity, leading in the end to the simple equation E=m*c^2.
Once this library of simple heuristics is large enough, they can be combined into the next Einstein.

When will we have it? I don’t know, I am not Einstein.