Monday, March 20, 2017

Cooldimi solution for Customer Experience

Kuldip Pabla is the co-author of this blog

CX (Customers Experience) 

We think we know what it is, at least it has a definition in Wikipedia:
In commerce, customer experience (CX) is the product of an interaction between an organization and a customer over the duration of their relationship. This interaction is made up of three parts: the customer journey, the brand touchpoints the customer interacts with, and the environments the customer experiences (including digital environment) during their experience.
A good customer experience means that the individual's experience during all points of contact matches the individual's expectations. 
A Customer Experience is part of the almost infinite number of human experiences. A negative CX creates pain in customers. Here is a tweet

CX is broken across industries

The article 5 Companies People Hate Calling  features Comcast, ATT, Wells Fargo, Time Warner Cable and Dish Network. We read

  • If you want an exercise in banking futility, call Wells Fargo. 
  • ATT had the dubious honor of ranking number one among all companies in a survey asking subscribers which companies have the worst customer service in the country.
You can read what the article says about the other companies. The Customer Experience in these companies is a journey through anger and pain.

Top management is aware.  Their real intention is to "manage the customer experience" (as Gartner describes the process).

At  Comcast, after One Year in to Comcast’s Massive Customer Service Overhaul  there is no real improvement. 

The scores are still low. And, Comcast’s gains might be due to their initiatives or it might be attributed to the Hawthorne Effect, where the mere act of looking at something more closely makes it improve.
Comcast's treatment for a  low CX was misdiagnosed. They did not develop a customer-focused vision and there is no awareness of the new Artificial Intelligence and Chatbots technology

Improving the CX  and eliminating customer pain

This is exactly what Cooldimi does

The following diagram shows the details

How Cooldimi Platform improves the CX
Most call centers assign one customer representative per customer, as you see in the top part of the diagram. If we have three customers calling at the same time, customer number four will wait. The wait times are longer and longer as traffic increases. Once the customer reaches the first level CR,  the person just starts collecting basic information and decide to pass to the next level, who may be not (and often she is not) the right person. There is enough wait time accumulating to breed anger and frustration

The lower part of the diagram shows how AI chatbots work in customer interaction. First, the chatbot responds immediately, there is no waiting time. If properly trained, the AI chatbot handles the salutation, standard small talk and collects all the data.

The Cooldimi platform uses chatbots to do the heavy lifting. One human CR can service now 10x or more customers and keep them happy

Research has shown nearly half of people do not mind chatting with chatbot, as long as the conversation is taken over by a human later on. Each interaction of the bot uses the Knowledge Base of answers and at the same time updates it real with latest information

Sentiment Analysis

Monitoring customer satisfaction real time
Cooldimi Platform displays the customer satisfaction. In previous blogs, we talked about happiness factor. In large companies there is no human contact face to face among employees, management and customers. There is no personalization and this goes against human nature.

The diagram shows a customer, well identified, starts scoring low and falls in the red area;  the company takes immediate action and  brings her back to the yellow area

This is how Cooldimi manages the CX (Gartner's terminology). CX is an asset making money and improving profitability

Changing the culture

This can not be done overnight and it can not be done, from inside the company 
Changing an organization’s culture is one of the most difficult leadership challenges. That’s because an organization’s culture comprises an interlocking set of goals, roles, processes, values, communications practices, attitudes and assumptions.
The elements fit together as an mutually reinforcing system and combine to prevent any attempt to change it.
What we say  we do is visible. What we really do is based on inertia

It has to be done from outside. AI Bots for big businesses for Customer Experience are almost exclusively offered by startups  

We can help  turn a big ship around. We plant a nucleus based on AI and grow it up until a new standard of CX  permeates the organization. Changing the CX is the first step.

To avoid customer switching brands, in 2016 almost 90%  of companies compete based on CX

Saturday, March 04, 2017

Why Joaquin Quiñonero Candela left Spain ?

European ideas of creating wealth

"With all my respect for startups and dotcoms, they do not manufacture anything"

I took this quote and photo from a tweet on the book Leadership made in Spain . It has 12 interviews of entrepreneurs from Spain, one of them named Espinosa de los Monteros.

In twenty-five years during the 13th century, the Mongol army subjugated more lands and people than the Romans had conquered in four hundred years. They never manufactured anything. Their wealth was growing through war and pillaging.

In middle ages, wealth meant having land. Manufacturing goods exploded in England in the 19th century

There many ways of creating wealth, but the type of wealth created in Silicon Valley is least understood in the outside world.

Take for example Snapchat. After the IPO last week, the company, a text messenger in essence, is valued 24 billions, more than Best Buy or United Airlines. Snapchat lost $500 million and they will continue to loose money. How is this possible?

How Facebook creates wealth?

Then I read the article Inside Facebook AI Machine  Quotes
Facebook today cannot exist without AI. Every time you use Facebook or Instagram or Messenger, you may not realize it, but your experiences are being powered by AI.
We can not manufacture AI or ML. But Europe knows this very well. Here is the head of Artificial Intelligence in Facebook

 He comes from Spain. His name is Joaquin Quiñonero Candela. He is one of the many brilliant scientists from Europe.
Candela was born in Spain. His family moved to Morocco when he was three, and he attended French language schools there. Though his grades were equally high in science and humanities, he decided to attend college in Madrid, ideally studying the hardest subject he could think of: telecommunications engineering, which not only required a mastery of physical stuff like antennas and amplifiers, but also an understanding of data, which was “really cool.” He fell under the spell of a professor who proselytized adaptive systems. 

 Why brilliant minds leave Europe and Spain in particular?

Because the entrepreneurial elite of Spain did not keep the pace with the new ideas of wealth creation. The quote by Espinosa de los Monteros. at the beginning of this post, makes impossible for scientists like Candela to create a similar wealth there.

At the end I see this picture from Facebook Building 20 in Menlo Park. This is where I worked when Sun Microsystems still existed

And this only tiny part of Facebook AI operations.


We are cooldimi . We are a startup. Of course we are not Facebook (not many companies are). But we create wealth via AI and ML for companies who are not the size of Facebook, and whose business focus is different.

We can monetize hidden treasures all companies have. We increase the CX (customer experience)

Friday, February 24, 2017

Cooldimi Chabad Bot is alive

What is Chabad?

Chabad is considered one the best organized Chassidic Jewish organization in the world They have a web site with over hundred thousands pages.

Today, over 3,500 Chabad institutions are located in more than 85 countries, with a new center opening on the average every ten days. In South Africa, South America, Russia, Australia, the UK, and many parts of the USA,

Chabad JCC from Placer County

This is one of the 1.500 Chabad Houses worldwide.

The Chabad Jewish Community Center from Placer County  has it's own web site. Since February 22, 2017 you will see a  "Message Us" Icon in the upper left corner. The link leads you to an Artificial Intelligence Cooldimi bot,

What the Bot Does

Placer County JCC, in California is the first ever to implement an artificial intelligence bot.

It answers relevant simple  questions like these

Even if the question is misspelled, the bot understands

When the Shabbat starts? Note that each of the world locations has a different time.

In case you want to know how Chabad started in Placer County, the Bot has an answer

One must-have feature is that Chabad Bot shuts down during Shabbat and starts again answering only after the Shabbat ends

Anyone answering a question, as a member of Facebook community, can be identified. A new feature to be released soon, the Rabbi can broadcast text messages via the bot, only to people who need to have a conversation.

Note how we reduced the domain of the bot only to knowledge useful for people in Placer County and visitors. Otherwise the facts containing Judaism and  Chassidut are impossible to manage by a single bot.

The empathy

The Placer County JCC bot does not take any job away from anyone. There is no budget  to have human employee or an answering service replacing the "Message Us" bot.

The Chabad philosophy is to look past the outward person and trust in the soul deep inside. You may disapprove of everything he does, and his outlook on life may be the opposite of everything you believe. Don’t argue with him. Instead, be one with him. Unconditionally.

Ultimately the Rabbi Yosi Korik of  is the one  human the bot refers to when it runs out of answers. It frees the Rabbi to reach more people in need of more answers

The more people use the bot, the better will get via machine learning. Right now it gets nearly 70% accuracy and we hope to bring it to 90% accuracy.

But one thing the bot will never replace is human empathy. This is the main ingredient of Customer Experience (CX).

The wisdom from the creation of the Chabad bot applies also to a Cooldimi prototype bot selling car insurance. About this, we will talk in a future blog

Wednesday, February 08, 2017

Who teaches machines how to learn?

From the previous blog
 The ML model at the right  becomes useful, when we have lots of data fed continuously,  Training the ML model takes up over 80% of the time required to a useful Artificial Intelligent model, good enough, but never never absolutely perfect. Without data, ML does not work

Who is the teacher of a machine learning?

Classical idea of a teacher is the one who knows everything telling the pupil who knows nothing how to do things. This idea is medieval,
Confucius brought light
Do I regard myself as a possessor of wisdom? Far from it?But if even a simple peasant comes in all sincerity and asks me a question, I am ready to thrash the matter out, with all its pros and cons, to the very end.
Teaching in Confucian definition is sharing  doubts and experiences with other intelligent people in a process that leads to new knowledge. Note that a machine is not human, it is a dummy

 Teaching is the best way to Learn

This is again and again a re-discovered truth by creative people.  It is also a title of a Medium blog of Joe Toscano, the Google star CX UX designer. He describes his experience
 Teaching will make you think different
 Teaching will humble you, when one see getting across to people
 Teaching is the best way to learn. If there’s one thing I recommend to all professionals, it’s that you get out and teach.

The Ultimate guide to Chatbots

In this blog of Joe Toscano _- published in October 2016 – it is too early to write about Bot Building Best Practices. Each thought has value, but it does not form something unified, not even remotely.
His advise is:
  • Bots should not take over the role of a human being
  • Use your knowledge of user’s sentiments to trigger content blocks to extend and improve the experience at the most appropriate time.
  • Designing for human emotion is just as, if not more, important with chatbots as it is with any other experience.
  • Accept that building your AI is going to take time and that initially it’s not going to work well. The only way it gets better is when it breaks. Plan for this and create boundaries (narrow domains) to limit breaks.

In Cooldimi, we went beyond technological creation of bots

We go back to Confucius who lived two and half millenniums ago. It says that teaching is sharing  doubts and experiences with other intelligent people in a process.
Who is the teacher? More important than teacher, who is the Orchestra Director?
We can demo how we created our Cooldimi bot training pilot cockpit with all the tools required to orchestrate the machine learning. We go beyond technology We believe in an unbiased sentiment analysis interwoven in the bot texture.
Cooldimi believes in changes, in continuous content updates and improvement, and it can take care of it for your organization.
Like Confucius, we welcome anyone to ask a question. This question will lead to the knowledge you already have hidden but you did not know how to extract it yet yet.

Cooldimi way to orchestrate initial content

  • Make it conversational – conversation could be back and forth a few times.
  • Think from users point of view instead of enterprise’s. What you want to tell them?
  • Think of multiple conversations – even though there is duplication of information, that should be fine.
  • Always add greetings and thanks
  • Don’t add seasonal, temporary information, unless someone will update it regularly and timely
  • Define clear alarms to trigger human intervention
  • Compare against the goals of the bot all the time

Thursday, January 19, 2017

AI for Dummies

Artificial Intelligence for Dummies

Only in US, a dummy means a stupid person

In UK according  Oxford Dictionary, a dummy means a model or replica of a human being or an object designed to resemble and serve as a substitute for the real or usual one

Artificial intelligence is a dummy of the real Human Intelligence. From Quora 
  • Artificial means man-made. The only way we can be considered artificially intelligent is if we considered biological reproduction resulting in intelligent humans as a man-made process. 
  • We know that we are not artificial intelligence because we know that we are not artificial
So what makes the best Artificial Intelligence (AI)  is the human mind. AI is nothing but a tool used by creative people.

Machine Learning

This is a screen shot from the video Machine Learning Made Easy presented by Shashank Prassana from MatLab.

Figure 1.Standard Approach and Machine Learning Approach
As the definition says, Machine Learning (ML) uses data and produces a computer program to perform a task.

  1. The Standard Approach, on the left is to write by hand a program or use a formula to perform the task of human activity detection
  2. The right side is the ML Approach. We use a Machine Learning algorithm. There is no formula to detect human activity. If such a formula would have had existed, there is no need for Machine Learning. We consider Machine Learning for complex problems, where there is no magic equation.

The ML model at the right  becomes useful, when we have lots of data fed continuously,  Training the ML model takes up over 80% of the time required to a useful Artificial Intelligent model, good enough, but never never absolutely perfect

My  perception is most multinationals naively believe in magic suppliers who will offer off-the-shelf  bots they like and their customer will be delighted.

Cooldimi Platform

Cooldimi is a B2B and B2B2C Customer Engagement platform designed for the conversational economy, meaning one can order services, get important information and manage infrastructure without leaving the messenger chat.

Since October 2016, the mobile Internet usage exceeds laptops

The goal to design a chatbot undistinguished from a human, is unachievable. There is a moment when Artificial intelligence becomes Autistic-like intelligence.

Here is quote from The Register
“Alexa, are you the best chatbot in town?” “Sorry, I don’t I understand the question I heard,” she replies.
Cooldimi  blends Artificial Intelligence with Human Intelligence.We monitor the bot to make sure it delivers satisfaction to other humans. Second, Cooldimi defines a narrow domain personalized for the user.

The studies show, 44% of the users like talking to a chatbot to begin with. 88% of the users, expect a human intervention to follow

Tuesday, December 27, 2016

The disappearance of Big Data

Quoting from my blog April 2016

No one will be talking Big Data in 24 month (November 2017)

It  is going to get so good, so fast, so in 24 month, I mean literally in less than 24 months, that no one will be talking about Big Data. We will just spit in data like using a fire hose, and then spit out patterns predictions, correlations, causation, that we could never  understood. These technologies are compressing things so hard and our brains are built linearly. That why we can not see it.

We will not bother with Hadoop 3.0. We will use "this thing"

I do believe that the notion of Hadoop 3.0 would  simply be, "we will not even bother with it."  

Yes, indeed. It did already happen, December 2016

Sunday, December 18, 2016

The myth of Google Translate

The article The Great A.I. Awakening is typical of the flamboyant style journalists from New York Times use to create an artificial suspense ex-nihilo
Late one Friday night in early November, Jun Rekimoto, a distinguished professor of human-computer interaction at the University of Tokyo, was online preparing for a lecture when he began to notice some peculiar posts rolling in on social media. Apparently Google Translate, the company’s popular machine-translation service, had suddenly and almost immeasurably improved. Rekimoto visited Translate himself and began to experiment with it. He was astonished. He had to go to sleep, but Translate refused to relax its grip on his imagination.
The highlighted text  is added to manipulates the reader. Profesor Rekimoto, went on and tested a Japanese version of Hemingway's The Snows of Kilimanjaro translated  back into English using Google Translate. The result is decent.

This made me experiment too. As I was born in Romania, I selected the classic writer -  considered impossible to translate,  Ion Luca Caragiale.

Here is the Romanian version of a fragment extracted from the short sory The Visit
După un moment, se-ntoarce cu cheseaua goală; o pune pe o masă, se apropie de mine, îmi ia de pe mescioară tabacherea cu țigarete regale, scoate una, o pune în gură și mă salută militărește, ca orice soldat care cere unui țivil să-i împrumute foc. Eu nu știu ce trebuie să fac. Mama, râzând, îmi face cu ochiul și mă-ndeamnă să servesc pe domnul maior. Întind țigareta mea, militarul o aprinde pe a lui și, fumând, ca orice militar, se plimbă foarte grav de colo până colo. Eu nu-l pot admira îndestul, pe când mama îl scuipă, să nu-l deoache, și îmi zice:
— Scuipă-l, să nu mi-l deochi!
Here is the version from Google Translation
After a time, coming back with empty cheseaua; puts it on a table, get close to me, I take the small table royal Cigarette cigarette, take one, put it in his mouth and a military salute me, like any soldier who asked to borrow a ţivil fire. I do not know what to do. Mother, laughing, winking at me, and I-ndeamnă to serve Mr. Major. I extend my cigarette, lights it on his military and smoking, as any military walking from place to place very seriously. I do not admire îndestul, while mother spat, not to deoache, and I say:
- Spit it, not me him the evil eye!
Although Google did a decent translation, with only four words kept in  original Romanian, there is no sense of style. Some of the words are archaic and suggest a comic perfume the Romanian reader captures and laughs, while the English reader still struggles to make sense of the text and can't understand the reason why the Romanians are laughing.

So no mater how advanced the AI translation engine is and no matter if the Natural Language Understanding has a high rating, they can't translate from culture to culture the  sentiments which make literature a pleasure to read.

In my humble opinion, we can improve language , but we can not replace the people completely. We still need very skilled human translators to take over from AI/ML. The leading companies in applied AI  understand this very well.

The suspense from the New York Times article has been created by a  human,  GIDEON LEWIS-KRAUS,  and not by a machine

Love me some Google translate (Go here for the magic  ):

Monday, December 05, 2016

What makes Cooldimi bot different?

Written by Kuldip S. Pabla, co-founder and CTO of Cooldimi 

What is Cooldimi?

"Cooldimi is a stealth mode startup in based out of Silicon Valley. Cooldimi's bot platform is focused on enabling conversational commerce by using Chatbots. Cooldimi thinks that AI machine learning Chatbots will not replace human beings in the near future.

The  Yale University Professor David Gelenter's  view  is  The Danger is not Machines Becoming Humans, but Humans Becoming Machines

As a result, Cooldimi's unique approach involves creating super intelligent bots by blending artificial intelligence alongside human intelligence, which is irreplaceable in bots technology for the time being.

What is "bite-sized content"?

A bite-sized content is what a bite-sized snack is as compared to a normal sized meal. By analogy, the bite sized content came up with the advent of the millennials generation. They do not like long articles, long videos; they like shorter articles, shorter videos (5 minutes max versus 30 minutes).

One of the customers - a popular science center- shared that they have a website that they continuously update. However, the customer has experienced lower web traffic that has been  going down, though their site is continuously updated and has excellent contents. In my discussion, I suggested a messenger based interface, where the users can ask for smaller, shorter questions, and the responses are smaller, shorter and relevant answers.  The customer would love to use that service. More important for desktops, laptops, tablets you need to open the website. In Cooldimi solution,  all you need is to send a message, ask a question and then you get back the relevant responses. The customer marketing  representative also would like to know whether their marketing campaign are successful or not. In addition they liked this  idea, as is so focused to the point, because youngsters with a limited attention span would like to use messengers rather that browse entire websites looking  for a specific answer.

They want to understand what is the real meaning  of the  questions people have (Gap Analysis) - What people really mean? What questions they should really ask, and they don’t?

Why a Cooldimi bot is different?

Most of the bots - according to some sources they estimate close a million bots floating the Internet - are rule based. I would say between 80% to 90%  bots are rule based. They behave exactly how they are programmed to. They cannot learn and cannot respond to anything they do not know.

The next generation of bots is when a user can communicate in natural language, the way speak. This is also referred to as Natural User Interface (NUI), like GUI. So if I say for example: What is the price of apples? or How much the apples cost? , while the intention is the same, different users may ask it in many different ways. Using Natural Language Understanding (NLU), the second generation bots can understand intentions and can answer the queries even if they are not programmed to answer so. These bots can also be continuously trained to augment their smartness.

That said, these bots cannot do what a human can do - understand emotions and sentiments. Humans in a face to face interaction, can see the gestures, voice inflections, the body language. One can see whether the person I am talking to is happy or not, maybe she or he is frustrated or enthusiastic . These are important cues in sales interaction.

But when  talking to a broad audience of humans, how would an inanimate bot fine tune these conversations? It has to be done through sentiment analysis, and this is how a bot from cooldimi stands out. Not just we analyze sentiments, we use the cues to improve bot behavior to make it smarter.

Sunday, November 13, 2016

Apcera re-purposed itself.


Apcera started as a PaaS platform called Continuum; then HCOS (Hybrid Cloud Operating System); then simply Apcera  platform.

The latest Apcera incarnation from Network World Nov 8  is
 Apcera’s container management platform for cloud-native and legacy applications. At the same time, the company also announced it has expanded integration features to further support Amazon EC2 Container Service (Amazon ECS), Google Compute Engine and Microsoft Azure—strong integrations that should further the company's claim to providing near-seamless movement of workloads across heterogeneous environments.
According Mark Thiele, chief strategy officer for Apcera.
“Apcera has an enterprise-first approach to container management, and customers regularly turn to us to help modernize legacy applications, as well as to run Docker securely and at scale,”


There is a January 2015 - famous now - blog from Bryan Cantrill, Joyent CTO . The title is Predicteria 2015 . If you wonder what Predicteria means, Bryan told me and I wrote it down in this The new Joyent blog.

According to Bryan
  1. 2015 is the year of the container. ... Thanks to Docker, the world is finally figuring out that OS-based virtualization is actually highly disruptive (better performance and lower cost!), and I think that this realization will become mainstream in 2015. I don’t think that Docker will necessarily be the only substrate,
  2.  The impedance mismatch between containers in development and containers in production will be a wellspring of innovation. Currently, containers have a ton of developer enthusiasm — but limited production deployments, in part because of production concerns around security, persistence and network virtualization. But it’s a sure bet that there will be (many) players tackling the problems in interesting ways.
 I asked, how did  he notice the impedance mismatch?
When I speak to a developer event, I ask people the audience: "How many of you have used Docker containers?"  About one third to half of the audience raises the hand. Then I ask:  "How many of you have used Docker containers in production?" and I see just a few hands  still up
I am convinced it was the vision of Bryan that lead Joyent to re-invent itself

Why Docker is a winner versus VMs

This is the title of my LinkedIn blog . I am humbled by the traffic generated from Docker, Quora and LinkedIn

I ended the article like this
If Docker is a contract between devs and devops, then Apcera's HCOS seals it in form of ITOps policies that make it  respected and implemented in an automated fashion
I still believe this today.

Friday, November 04, 2016

How to stop AI bots from creating a Mad Max world

What Cooldimi offers?

We enable intelligent Chat-Robots, in short Bots. We make a bot smarter. To use an analogy, a hunter needs a dog. But this is not any dog, but a special breed and trained as a hunting dog.

Similar to a hunting dog, a cooldimi bot learns faster to make successful it's owner, knows it's  limitation and lets the human take initiative when it has no answer, It will quickly learn new skills to improve his owner satisfaction.

We have demo available  on how we actually do this with a bot, any bot from Facebook, Microsoft, Pandorabot and so on.

What can I do with a bot?

Before answering this question, Facebook’s introduction of a bot-building platform at its F8 developer conference event April 2016, triggered indirectly some colossal profit increase for the company. Facebook had a wildly successful Q3, earning $7.01 billion in revenue. Facebook destroyed analyst estimates, which were $6.92 billion in revenue 
The first thing you can do with a bot you can make money for your company.
This is Cooldimi's bot platform ultimate goal.

Do all bots have intelligence?


Do you mean machine learning intelligence?  No. Most of them are rules based. Like this Rubik's Cube Kuldip Pabla (Cooldimi CTO) built in one day for his son who trains for Rubik speed solving competitions.

So what is Cooldimi?

Cooldimi is bot platform.   It enables Conversational Economy, meaning one can order services and manage infrastructure with leaving the messenger chat.

It focuses on industry specific (narrow) domains, For example a bot calling an Uber car can not be used for email promotions

Cooldimi measures happiness factor using human sentiments It  blends Artificial Intelligence with Human Intelligence.

Cooldimi makes bots that meet business specific goals  and the users are happy. People never buy something they are not happy with.

Cooldimi adds a meaningful conversation in a suitable narrow domain. with clear business goals, directly or  indirectly, and we intelligently monitor the bot to make sure it delivers. Do you want to make money?

That will be nice.

Cooldimi makes the difference between (money-in to  garbage-out) versus (money-in to a-lot-more-money-out)

Can Artificial Intelligence appear as human intelligence?

Not  in the foreseeable feature

According to WILDML blog 
Many companies are hoping to develop bots to have natural conversations indistinguishable from human ones, and many are claiming to be using NLP (Natural Language Processing)  and Deep Learning techniques to make this possible. But with all the hype around AI it’s sometimes difficult to tell fact from fiction.
 Most of the value of deep learning today is in narrow domains where you can get a lot of data. Here’s one example of something it cannot do: have a meaningful conversation. There are demos, and if you cherry-pick the conversation, it looks like it’s having a meaningful conversation, but if you actually try it yourself, it quickly goes off the rails.

What do you mean "it quickly goes of the rails?"

If you’ve actually used any of the personal assistants (PSa) you may be skeptical. Siri still barely understands what you want, and Facebook has put hordes of human workers behind Facebook M to get it to do anything useful. How will these things ever replace all the complex tasks we’re doing apps and browsers? 
So you need human intervention. Machine learning can not sort it out by themselves, Have you heard Microsoft TAY?

I did. It made an uproar. How it happened?

It is a failure that was a success, as we learned so much from it. See Microsoft’s disastrous Tay experiment shows the hidden dangers of AI  
Microsoft’s disastrous chatbot Tay was meant to be a clever experiment in artificial intelligence and machine learning. The bot would speak like millennials, learning from the people it interacted with on Twitter and the messaging apps Kik and GroupMe. But it took less than 24 hours for Tay’s cheery greeting of “Humans are super cool!” to morph into the decidedly less bubbly “Hitler was right.”

How Coodimi avoids a Mad Max invasion?

The domain millennials mean all people aged 18 to 36 . This a huge land, impossible manage to via artificial intelligence, where one starts training a bot with an initial conversation, and improves this conversation as the bot gets used more and more. A bot has no notion of ethics. The users take control of the space creating out of Chaos an unpredictable world. Just like the movie Mad Max, a 1979 Australian post-apocalyptic dystopian action film.

24m of Twitter‘s 284m active users are bots with “no discernible user action involved”,  This was back in January 2015. Today the number can be much larger.

A clear domain narrowing is key for enabling  business bots that deliver results

Monday, October 10, 2016

Cooldimi Bots

Americans do not like long words. A Bot is a  shortening of  ChatRobot. A poet, Eli Siegel wrote
is what remains
when everything you think of
has gone

 2016 will be the year of conversational commerce

This is the Medium blog by Chris Messina
... utilizing chat, messaging, or other natural language interfaces (i.e. voice) to interact with people, brands, or services and bots that heretofore have had no real place in the bidirectional, asynchronous messaging context
The Aha! came from this observation


What does it mean? 

No longer do you need to convince users to “download and install” an app — they can just invite a bot to a conversation and interact with it [eventually] like they would a person. Zero barriers to adoption, with minimal risk to the user (i.e. malware, etc).
While you may have bristled when that news app alerted you to “new stories”, you might appreciate a particularly friendly newsbot delivering a personalized recommendation with context that you uniquely care about.

Bots have no menus and are no human either

You just text, or talk. "Conversational apps are therefore organized around the way I organize my life, rather than the way the app maker might dictate."

Bots are not humans. Bots will not become indistinguishable from humans in the foreseeable future. Bots have no sentiments

95% of the bots lack intelligence and are not learning from human interaction

Artificial Intelligence is hot

From 14 deals to 143 deals in five years

Bots themselves have not feelings. But the people using them have emotions.

The bot substance is..

To make happy the humans using them ." this is what remains when everything you think of has gone"

Cooldimi Bot

Diagram of Cooldimi Bot

Cooldimi rules

  1. Goal-driven dialogue bots that learn to engage in natural conversations with humans. It will recognize "Hi, how are you today?" or "It's a nice day in Arizona" as greetings. It will take more time to learn how to place the greetings at the right place
  2. We optimize users’ satisfaction and agent’s knowledge 
  3. We know that Bots are AI machines talking to real humans.A caller to some of the  government voice mail system is furious after waiting for hours. This can no happen with Cooldimi bot. 
  4. Cooldimi is enabler for a company to have happy users. Most other 3rd party are disablers.
  5. Bots are specific for each vertical. No mix and match, please, as a recent 3rd party discovered the hard way. A 40 verticals "ready to use bots" went out of business in 3 months
Cooldimi Bot: Demo will be available on web site during Nov 2016

3rd party disablers

A new start up in New York area uses peer too peer bots to sell insurance from a smart phone. Their business model requires the society to change. They even hired a Chief Behavioral Officer.

In my humble opinion any company that needs to change the people behavior, as a precondition for success has to fight two wars with very meager startup resources

From Talmud

Seek not things that are too hard for you. Search no things that are hidden from you. Think only at the things that are permitted to you. You have no business with the things that are secret.

Chatbots And The Future Of Facebook Messenger
Note: The ideas expressed in this article are  from Cooldimi  co-founders: Kuldip Pabla and Miha Ahronovitz

Tuesday, September 06, 2016

Insurance Industry and the next Industrial revolution

At World Economic Forum in Davos 2016, this picture feels grandiose

3 key questions for the digital revolution, by Jim Snabe, WEF USA
According to Harvard Business Review , the article Is Your Organization Ready for Total Digitization? explains:
What do the following items have in common: credit cards and streaming or recorded music, robots for production, CAD systems, telephone networks, digital games, computers in products like cars and vacuum cleaners, sensors, and video consoles used in remote mining? Answer: They are all digital and connectable.
This is the world of total digitization: a multitude of digital devices and sensors creating streams of data, as well as any number of digital services and products for both internal and external use, distributed throughout the enterprise, and sometimes, but not always, connected.
However the road to the 4th industrial has the devils in the details.

The amazing challenges of digitization the life insurance industry

This is an industry that bills total premiums of 2.3 trillions dollars per year worldwide As part of the work we did in Cooldimi we discovered  some interesting facts

Insurance documents are verbose and complex

On average insurance products come with "more than 25,669 words of explanation written in PhD level language." This is more than Shakespeare's Romeo and Juliet . See Customers ‘only read 15%' of insurance documents they receive' . One of the big enemies for getting the right cover is complexity.
There two startups trying to bring life insurance on line.   Quilt coming soon
"Quilt is for people who live online, value what they have, and expect more from their insurance. We promise: we'll never make you read pages of legalese or hound you with calls from salesy agents".
Another is PolicyGenius that compares life, renters, pet and health insurance on line. Their mere existence is a proof of Insurtech 3rd party disruption, I see why their methods can not be adapted gradually in all incsurance industry

No national databases of all life insurance policies

This is unthinkable
The following article has a self-explanatory title: 12 Easy Steps To Locating Lost Life Insurance Policy Documents. Quote:
Locating life insurance documents for a deceased relative can be a daunting task—for one thing, as of this moment there are no national databases of all life insurance policies.
There are reports of over 1 billion dollars in  unclaimed life insurance - accumulated over the years in US alone.

The role of human happiness in selling insurance products

 Why are people reluctant to buy life insurance?

From a millennial (someone in her mid 30's) Quora contributor
People are often reluctant to join a plan that provides no immediate benefit. Not only does life insurance not provide an immediate benefit, but it doesn’t provide any at all as long as you are alive. It’s likely that many people are turned off by the idea that they will not personally benefit from having a life insurance policy.
There are people who have an intense fear of their own death. It’s not that any of us are somehow okay with that prospect, but some people fear it to the degree that they don’t want to talk about it, and don’t want to plan for it either. Some might even believe that planning for their own death, such as buying life insurance, is like willing it to happen
Jewish orthodox people (and many other religions) believe our life is sustained by Gd and we could  trust  a life insurance policy the way we trust Gd.  According to Rabbi Alan Yuter
The head of the household viewed children as old age insurance; children were expected to care for the mother in old age. Life insurance fills this social need. The marriage document is at its core a life insurance policy which the wife collects at her husband’s demise or at her divorce, i.e., the demise of the marriage. 
For Hebrew culture, love requires tangible expression. This expression includes giving financial and emotional security for those who depend upon us.
These sentiments are attached to the product called life insurance.  Simply lowering the price of the product will not make it more desirable

Happiness  Index 

The 4th Industrial revolution will not happen, unless we digitize  the human sentiments in addition to credit cards, life insurance, and other soulless infrastructure objects.  Using machine learning and sentiment analysis, we should always have a clearer picture of the buyers and sellers happiness. See the blog  Happiness in Business

Measuring happiness is not an interview process. People are not entirely honest in interviews, and the samples are too small.

Cooldimi translates and monitors continuously the Sentiment Analysis from   business data into a  Happiness Index , per product, line of business, geography or society impact. We detect the polarity (negative, neutral, or positive sentiments) in real time using Natural Language Processing (NLP).

Otherwise the barriers between man and machine won't dissolve.

Post Script

Amazon started using machine learning in consumer products reviews evaluation. Just click the stars in any review like this, and this text pops up.
Amazon calculates a product’s star ratings using a machine learned model instead of a raw data average. The machine learned model takes into account factors including: the age of a review, helpfulness votes by customers and whether the reviews are from verified purchases
If Amazon stopped calculating simply raw averages in rating, everyone will follow soon. Behind each human rating, there is a sentiment: joy, envy, maliciousness, sincerity, fanaticism and so on

Thursday, August 18, 2016

Artificial Intelligence or Artificial Narrowness ?

Artificial Intelligence (AI) is “intelligence” that is not the result of human cogitation.

I followed the question on Quora What is the difference between artificial intelligence and human intelligence?  and here are some answers from clever contributors
"Artificial means man-made. The only way we can be considered artificially intelligent is if we considered biological reproduction resulting in intelligent humans as a man-made process."
"We know that we are not artificial intelligence because we know that we are not artificial."
"In my opinion it is misleading to talk about "artificial intelligence". The alternative term, "machine intelligence", is far better."
Machine Learning (ML) is the subset of the Artificial Intelligence (AI)  focused on algorithmic approaches, while traditional methods mostly use a parametric approach.  Yet ML and AI are used almost interchangeably


Artificial intelligence usefulness is not apparent to every one

Lets assume an author who is adapting complex ideas for public consumption must be a.) capable of understanding the ideas thoroughly in their original form; and b.) familiar with the limits of understanding of the general public and the way these materials must be presented to them.

We do not make money as a business organization by cutting customer service jobs and replacing people with bots. We make money through people
Machine learning is here to give the management a view of profits, along side the human assets sentiments who are the true generators of wealth in well managed companies.

One can not have ultimately a company made out of token activated bots.

Imaginative uses of ML

The article why Good Companies Go Bad from Harvard Business Review 1999, describes the term Active Inertia. This is an organization’s tendency to follow established patterns of behavior—even in response to dramatic environmental shifts. Many large companies fall victims. Machine Learning may have access to documents where it can extract proof of the Active Inertia. Than the management makes changes and see if the symptoms of Active Inertia are less than before.

The happiness literature has identified one of the most deeply satisfying human psychological states to be one called “flow.” It occurs when you are so immersed in an activity that you lose track of the passage of time. Can we identify roughly what is the the percentage of total employees who experienced it?

Finally, we want in our country to give people the ability to work productively in steady jobs, without the stressful fear of being riffed. Most humans are not entrepreneurs and are risk averse.

Look at Intel, Cisco, and the like laying off people in massive numbers. They could have had avoided it is they have had identified the Active Inertia in their companies and the lack of flow in their employees.

Thursday, July 28, 2016

Happiness in Business

I will start by making the following statement:
Happiness is the biggest unused revenue making asset in a business organization today
In real life, when we meet socially a person, we have a conversation. After a few meetings and conversations, we unconsciously evaluate each other. We start having fuzzy opinions about what the other person needs, what she likes and what she dislikes.

A character in a hit Israeli TV series, is a young  medical doctor listed number four on a local list of most desirable bachelors. He is attracted by a young woman, Tehila, but he learns she only goes out with poets. This is an essential piece of information, specific, individualized  and can not be extracted easily through classic predictive analysis.

The happiness question in business

How can we create a tool to deliver this more precise information from interactions with millions of people? Could it make us trust each other more and replace the fear of rejection?  This applies equally to the employees of a company and it's prospects and customers.

The banner "we are nice people to do business with" if it gets across, can make tens, hundreds an even thousand more revenues as unprecedented growth

The advent of artificial intelligence (AI)

Ray Kurzweil,  futurist and inventor, now the director of engineering at Google believes human level Artificial Intelligence (AI) will be achieved by 2029

David Gerlerner  disagrees. The encyclopedic computer scientist whom Bill Joy, the co-founder of Sun Microsystems called "one of the most brilliant and visionary computer scientist of our time"  said you can not separate the mind from a body
The human mind is not just a creation of thoughts and data; it is also a product of feelings. The mind emerges from a particular person's experience of sensations, images and ideas. The mind is in a particular body. Engineers can build sophisticated robots, but they can not build human bodies

The Master Algorithm. Machine Learning

 This a revolutionary book by Pedro Domingos,   Tech Insider  makes the introduction
We're in the middle of a historic moment. It used to be the case that you had to program a computer so that it knew how to do things. Now computers can learn from experience.
The breakthrough is called "machine learning." It's unimaginably important for understanding where technology is going, and where society is going with it. 
Pedro Domingos says more:
A self-driving car is not programmed to drive itself. Nobody actually knows how to program a car to drive. We know how to drive, but we can’t even explain it to ourselves. The Google car learned by driving millions of miles, and observing people driving.
This is what happens with every single human driving a car. But we can not experience what millions other people do when driving their cars.
That's the key: machine learning allows algorithms to learn through experience, and do things we don't know how to make programs for. 
 There were two stages to the information age, One stage is where we had to program computers, and the second stage, which is now beginning, is where computers can program themselves by looking at data.
Google's Eric Schmidt says that every big startup over the next five years will have one thing in common: machine learning.

All models are wrong, but some are useful

On page 290, at the end of his book, Pedro Domingos writes
So now you know the secrets of machine learning. The engine that turns data into knowledge is no longer a black box: you know how the magic happens and what it can and can’t do.
 Here is another attempt of finding a spouse using Machine Learning
One hardy geek extracted twenty thousand profiles from OkCupid (dating online) , did his own data mining, found the woman of his dreams on the eighty-eighth date, and told his odyssey to Wired magazine. To succeed with fewer dates and less work, your two main tools are your profile and your responses to suggested matches
We need to produce  a learning algorithm that works. Using the wrong algorithm, it does not work. We have to know what we are doing,  and watch out  that we do not get garbage-in, garbage-out. I dug out the article from Wired on the web How a Math Genius Hacked OkCupid to Find True Love. You can read the details, the math. In my view, it is a classical example of garbage-in, garbage out for most people, except mathematician Chris McKinlay.

The goal of ML is never to make “perfect” guesses… The goal is to make guesses that are good enough to be useful. Because, as the British mathematician George E. P. Box  says: "All models are wrong, but some are useful."

Happiness makes the most successful businesses thrive even more

Happiness in Business

In his 2008 book The Geography of Bliss , Eric Weiner says
Happiness is a Number
He visits Rotterdam, Holland, where the World Happiness Database is maintained. Students at Claremont University in California can earn a Ph.D. in positive psychology- in happiness.

Since 2012, we have an annual  World Happiness Report Update  Denmark is the number one happiest country in the world for 2016.

Other countries close to the top are  Switzerland, closely followed by Iceland, Norway and Finland, What strikes me is these are relatively small counties in terms of population where people communicate and know each other best.

The report says:
Happiness is a better measure of human welfare than measuring education, health, poverty, income and good government separately, the report's editors argue.
There are at least seven key ingredients of happiness: People who live in the happiest countries have longer life expectancies, have more social support, have more freedom to make life choices, have lower perceptions of corruption, experience more generosity, experience less inequality of happiness and have a higher gross domestic product per capita, the report shows.
It is now the time to publish The World Happiness Report for business. Multinational companies like Ericsson, Zurich Insurance, Facebook, Microsoft, Shell have customers and do business all over the world. All we have now in use are Fortune lists (500, 2000 global, and the wealthiest people in the world)

The only way to write such reports is to implement using finesse the AI technologies,

Example from Insurance Industries

I will cover this in depth in a future blog. Based on my research, traditional insurance companies, billing worldwide 4.1 trillion dollar a year in premiums, are under attack by third party startups using artificial intelligence. Executive earning as much as 350,000 a year are fearing for their future jobs. Some stats predict 25% of the jobs in financial and insurance will be replaced by machine learning tasks. The insurance customers are millennials gen x people , under 50 year old. Insurance companies sell insurance essentially  the same way as 80 years ago, with few changes.

According to I2C: Insurance2Customer USA to be held September 19-20 in Chicago these are the new goals.
  • The Undisputed Rise of the Insurance Customer  – Manage the shift from product-centric to a customer-centric culture to meet sky-high customer demands
  • Understanding Your Customer to Build Relationships – Actionable insights to engage customers through targeted segmentation and personalized messaging
  • Creating a Memorable Insurance Customer Experience – Create and deliver a streamlined and simple customer experience across all channels
  • The Next Wave of Marketing Strategies – From social media, content marketing, branding and digital advertising, learn new tools to win and retain more customers


I hope this article touches a sensitive chord in the readers. Cooldimi is a stealth company. It's mission is to help companies, particularly insurance companies to set up meaningful machine learning projects

Reaching the customer insurance goals in 2016 requires machine learning implemented and interpreted with elegance and intelligence. Simply hiring a famous mathematician does not make companies reach the goals.

Because the goal is still to keep the executives employed and create bridges to happier and happier customers. Those companies will have happier employees, who can be promoted, not laid off. Will make easier for them to feel the flow. Robert Frank says in New York Times what the flow is:
The happiness literature has identified one of the most deeply satisfying human psychological states to be one called “flow.” It occurs when you are so immersed in an activity that you lose track of the passage of time. If you can land a job that enables you to experience substantial periods of flow, you will be among the most fortunate people on the planet. What’s more, as the years pass, you will almost surely develop deep expertise at whatever it is you’ve been doing.
A business that goes ten times in revenues has no need of laying off people,. Quite the contrary, it creates more jobs and can be mutually beneficial co-opetition. It will attract young customers and make the existing customers buy more.

Post Script: The test of happiness

The  Hungarian psychologist (with a hard to pronounce name)  Mihaly Csikszentmihalyi, also one of the founders of positive psychology. He introduced the concept of "flow" in 1975 He started studying creative people. This Pursuit of Happiness  web page has the mission of bringing the science of happiness to life. It took 40 years to bring the relatively dry, academic research in mainstream media. 

You can try to take a very simple happiness test on line.

The next step could be a  breakthrough article in Harvard Business Review (and other business prestigious publications) to make happiness widely adopted as business goal  Now we have the tools and the wisdom to make it practical

Blog Archive

About Me

My photo

AI and ML for Conversational Economy