More Soul, More Youthful Thinking and More Thinking Among Machines

brain-100691512-large

What is artificial intuition?

How can it be developed?

What if machines not only learn like children but also think like children?

What would happen if machines started to think together?

Bill Gates has suggested that robots should be taxed and that the money should be used to pay the people who have lost their jobs to robots. On the one hand, it makes sense to suggest that if robots will be depriving humans of work, then the company should simply pay tax for using their skills, and the money should be put into supporting the rest of society.  He also believes that certain jobs cannot be replaced by robots such as nurses and teachers.

Yet, perhaps we are just simply missing the point with using AI – Artifcial Intelligence. Perhaps AI offers a spark to kickstart a new way of building a society. A new way to ensure that everyone has a roof over their head and food on the table. New ideas are needed rather a simple application of the old practices.

The same thing could perhaps also be said about the way we think about machines and the way we design robots. If we look at perhaps the most transforming part of human history it is that we are not relying on individual thinking. Instead, the collaborative and collective thinking is one of driving forces behind our remarkable progress.

So perhaps we should focus on what potential there are among machines rather than within machines.

Moreover, the focus is often on building machines that can deal with increasingly higher volumes of data. Yet, to explore ideas such as building artificial intuition, may require that we instead look into ways that machines that use as little data as possible. Thin-slicing is a powerful concept. Designing machines that can improvise, without a script or a plot and react to new environments require new ways to approach the way we think about AI.

What if several machines could be connected to work intuitively on little information? Perhaps a solution could not be found by using this approach but maybe new insights and ways to approach a problem would emerge.

Children are the best learners. Developmental cognitive scientists and computer scientists have been working together to figure out how young children can learn so much so quickly. A problem with AI is that it has been very difficult to predict what aspects that would be most difficult to solve. Problems such as how to play chess and to detect statistical pattern have turned out to be fairly easy task to solve – admittedly. they could still be improved upon. Yet, a limited generalise can only be achieved from statistical learning, this is regardless of whether you are a child, an adult or a computer.

Children are often good at inventing new concepts and often their thinking is non-conventional – out-of-the box thinking. They link ideas and say things that do not make sense. Creating machines that could create new concepts and explore hypotheses that are not obvious could, just like listening to children, result in new insights.

What if you could transform the way we build AI? What would you do?

(Suggestion, read our other posts about intuition…..)

connectedmachines2

Save

Intensive Human Farming – Thinkibility Nibble

The success of industrial food production has been obtained by applying principles of scientific management, a theory of management that analyzes and synthesizes workflows. Its main objective is improving economic efficiency, especially labour productivity. It was one of the earliest attempts to apply science to management,and when applied to food production it has resulted in intensive animal farming.

intensieve veehouderij

Intensive animal farming or industrial livestock production, also called factory farming by opponents, is a modern form of intensive farming that refers to the keeping of livestock, such as cattle, poultry (including in “Battery cages”) and fish at higher stocking densities than is usually the case with other forms of animal agriculture — a practice typical in industrial farming by agribusinesses.

There are issues regarding whether factory farming is sustainable and ethical. From 2011 to 2014 each year between 15,000 and 30,000 people gathered under the theme We are fed up! in Berlin to protest against industrial livestock production.

816289-002

It is ironical to note that with the successes in the food industry at every turn, companies applied the successful principles of intensive animal farming (read: factory farming) for the management of their offices. Office management is the/an administrative handling, controlling, and maintaining a balanced process of work inside the office of an organization Topics in office management are Budget development and implementation, Purchasing, Book Keeping, Human resources, Accounting, Printing, Records management, Forms management, Payroll, Facilities management and Space management.

offices

If we nowadays look at large companies we could say that it is basically Intensive Human farming*. As are schools, hospitals and homes for the elderly.

The positive side is that already all steps have been taken to fully automate office management. But, what will we do when large sections of the population are unemployable through no fault of their own? What to do in the future where for most jobs “humans do not need to apply”? That is, only qualified robots are welcome. . . and needed. . .

*This metaphor was first launched by Jaap Peters and Judith Pouw in their book: De Intensieve Menshouderij

Key Performance (mis) Indicators

kpi

Key Performance Indicators are meant to keep an organisation on track. By measuring the performance over time, you are able to look at deviations and to take measures. As Wikipedia defines it: A  key performance indicator (KPI) is a type of performance measurement. An organization may use KPIs to evaluate its success, or to evaluate the success of a particular activity in which it is engaged. Sometimes success is defined in terms of making progress toward strategic goals, but often success is simply the repeated, periodic achievement of some level of operational goal (e.g. zero defects, 10/10 customer satisfaction, etc.).

The concept behind Key Performance Indicators is to build a feedback loop between input and output. Its working principle does not differ from a thermostat, which senses the temperature of a system so that the system’s temperature is maintained near a desired set-point.

500px-Feedback_loop_with_descriptions_svg

In order to get not market driven organisations more efficient the adage “The numbers tell the tale”has become fashion among governments, institutions and not-for-profit companies. There are several metrics or key performance indicators.

However, Key Performance Indicators can also lead to perverse incentives and unintended consequences as a result of employees working to the specific measurements at the expense of the actual quality or value of their work. In the social sciencesunintended consequences (sometimes unanticipated consequences or unforeseen consequences) are outcomes that are not the ones intended by a purposeful action. Perverse incentives are a type of unintended consequence. A perverse incentive is an incentive that has an unintended and undesirable result which is contrary to the interests of the incentive makers.

There are a lot of examples of bad designed Key Performance Indicators. We came across, but not exhaustive:

  • Police officers get a predetermined quota of fines to give out. The unintended effect of this KPI that the police organisation will be focused on easy to obtain files, f.i. traffic fines instead of fighting serious crime;
  • An organisation involved in handling objections has a KPI for the amount of rejected complaints. Imagine how employees will approach complaints. . .
  • It is generally accepted that the progress of students is evaluated by tests. However, student tests assess only a small part of needed knowledge, skills and attitude of students. Also, often the purpose of the test, timely warning of learning difficulties and study delays, dilutes to “a (missed) ticket to the next hurdle”;
  • An agency of child protection is responsible for placing abused or emotional neglected children in foster parents and child care institutions. It is very logic to design a KPI: like the number of placed children. If this performance is coupled to the financing of the agency, it can easily lead  to placing children out of their home, against sound indications that there is no need for or against parents objections;
  • It is complete reasonable to expect higher efficiency and experience of surgeons as a hospital performs at least 30 knee surgery or angioplasty a year. However, such a KPI can lead to more instead of less knee surgery and angioplasty, an example of a perverse effect contrary to what was originally intended (an intended solution makes a problem worse);
  •  The selling of mortgages as an end in itself, even to people who could no pay the interest, led to the bank crisis in 2008. Another example of a negative, unexpected detriment occurring in addition to the desired effect of the policy to motivate sellers to do better their best.
  • To increase the efficiency of university studies universities are judged on the number of successful students per year. It is now tempting to reduce the requirements for passing exams.
  • In order to increase the efficiency of General Practitioners many assurance companies allow for not more than ten minutes consults by patients. This KPI leads to far more referrals to medical specialists because GP’s have not much time to carefully investigate the medical complaints. This is an example of a counterproductive KPI: it is more of an “obstacle” than a help in the achieving of a productive project or an objective;
  • Crews of warships run annual series of nautical and operational exercises. Through a complex multi-factor analysis, a KPI is derived: Operational Employ-ability. Members of Parliaments asked questions when the KPI decreased to 10%, as a warship was actually deployed in a crisis;
  • Notorious are budgets: the setting of expenditure levels for each of an organization’s functions. It expresses strategic plans of business units, organizations, activities or events in measurable terms. However, such budget tends to be exhausted at the end of the year, because organizational units realise that they will be shortened in budget for next year, because last year they needed not the full budget. So, as an example, in many towns you can observe that every five to ten years the same streets and squares are completely overhauled without any need but in order to use the full budget.

Many Key Performance Indicators have unintended effects. They function as rules for behavior. Key performance Indicators are designed to notice need for adjustments of the course of an organisation. However, more often than not, they are invitations to cheat, by employees but equally by companies and institutions,  especially when financial consequences are attached to the KPI.

Whenever designing or encountering a Key Performance Indication, be warned!

 

unintended consequences

For more examples of perverse incentives, see here. For examples of unintended consequences see here.

To built up your Thinkibility skills, imagine your are the director of a hospice. You have set a thinking task: how to improve the occupancy (KPI) of the hospice. Then check your answers with How Dying Became A Multibillion-Dollar Industry.

 

 

My Portfolio – Thinkibility Nibble

ID-10041781

Artificial intelligence is spreading its wings and artificial identity and personality is here. Your Tweets may contain advertisements that are generated by keywords used in the Tweet. You may not be thrilled about the idea of artificial identities and resumes but not exploring the topic is not the solution.

Sures Kumar has created Pro-Folio,  the ultimate tool in creative plagiarism, to raise our awareness of artificial identities. The core algorithm can generate 690,903,803 trillion unique fictional identities! And it does not take long to create a stunning art portfolio.

A test showed that it took about 10 minutes on average to spot that a portfolio was fake. So there might be a market for software that detects fake portfolios! And that prevents material from being collected in a couple of seconds from website all over the world.

pro-folio_001

There is a disclaimer on the website and no harm is intended by using the artwork. The aim is to raise awareness.

Disclaimer: Pro-Folio.org is a HOAX project created as a part of a Scientific Hoax project at the Royal College of Art, London. More background information, context, briefing can be found in the Science Hoax project blog. All contents displayed in the website (http://www.pro-folio.org) are dynamically generated using server side scripts. The art and design works are fetched from various online sources and the designer’s name and details are generated from a wide variety of online databases. This is a speculative academic project aimed at intellectual stimulation and debate regarding the identities generated online and is not intended to hurt any individual or corporation.

Nevertheless, some people may react strongly towards Sures’ approach. What is your reaction? Is this a great way to make us think?

Synthetic identity theft is a serious crime and something that is becoming more prevalent. This identity consists of a combination of real security numbers mixed with a name and birthday other than the ones linked to the number. Why would you use a fake identity? Well, these synthetic identities are difficult to track and they do not appear on either person’s credit reports. Often a new file is created and this crime harms primarily the creditors who may give a person credit. One way to prevent crime is to think ahead and explore ideas that thieves might use. This approach has proven to be very difficult when you try to prevent online crimes, where the criminals have tended to be one step ahead.

Go here to create your own fake portfolio.

A link to Sures Kumar’s Portfolio.

Photo “Brain In Electric Bulb” by digitalart

Robots, Evolution and Emotions

ID-10013918

Can a robot evolve? Or is it chained to the insides of a program?

Names like Incher, Jitter and Wings, hit that the robots may not be simply ordinary robots.  The Cornell Creative Machines Lab has designed a program within which simulated robots “build” themselves out of cubes of virtual muscles and bones. They provided the computer program with  different materials and one rule. The materials resembled the basic components of our own bodies: bone, soft tissue and a couple of muscle. These flexible robots have developed unique gaits.

The rule is simple:

  • Robots that move faster get to reproduce more.

Over the course of 1,000 generations, you can see some amazing figures that flap and jump at various speeds.

So what is the point with making a programme like this? Well, the output from the programme was compared to scientists who were asked to design a better robot from scratch using the same parameters that the computer used. And the humans failed to produce soft robots like the ones that evolved in the computer simulation.

A combination of using soft tissues and evolutionary principles can help to design complex and interesting artificial life forms. Go here to read about how to use nature as inspiration for ideas.

00ac

Steven Spielberg’s film A.I.  has among several other films and books challenged us to think about the nature of human empathy when applied to a non-living thing such as a robot. And several studies have shown that children and adults can and will form emotional attachments to robots.

Using fMRI scans, researchers discovered that humans have emotional responses to how robots are treated. Based on the fMRI scans, the participants’ emotional responses to the treatment of humans closely mirrored their reactions to the good or bad treatment of the robot.

Astrid Rosenthal-von der Pütten said:

“One goal of current robotics research is to develop robotic companions that establish a long-term relationship with a human user, because robot companions can be useful and beneficial tools. They could assist elderly people in daily tasks and enable them to live longer autonomously in their homes, help disabled people in their environments, or keep patients engaged during the rehabilitation process.”

So what do you feel when you watch the video with the dinosaur robots used in the study?

Photo: “Hug” by graur codrin

World Thinkers’ Ideas – Creative Machines

Robots are machines that are programmed to perform tasks. Can a robot be creative? And how can you use robots as inspiration for new insights.

Driven by a desire to build a scientist smarter than himself, Jürgen Schmidhuber decided to become an artificial intelligence expert.  He believes that our dominated place as creativity experts may end around 2040 when Omega, or Singularity, will mean the end of our dominated position.

What is this idea based on?

Predicting the future is tricky, and although science fiction can provide great inspiration, few things can be predicted confidently. But that there will be computers faster than the human brain may be one of those predictions. Jürgen Schmidhuber may sound too optimistic when he says that computers will also be able to solve problems faster than humans can. However, at the Swiss AI Lab IDSIA research into artificial neural networks and reinforcement learning means that the day when machines are faster than us is creeping closer.

Can you read French, Arabic or Chinese handwriting?

Machines can and the fascinating thing is that it is not  a program. Instead, the machine has learnt from extracting regularities and making generalisations from data.

The step to being creative may look enormous. However, the project Formal Theory of Fun and Creativity may prove that it is possible. A theory has been developed that explains in a formal way science, art, music, and humour. Building curious and creative agents that never stop generating new ideas may be a work in progress, but when computer have the power of human brains, the explosion in ideas is a reality. He says, curiosity is the desire to create or discover more non-random, non-arbitrary, regular data that is novel and surprising.

Does the idea sound scary? Jürgen Schmidhuber does not think in terms humans versus robots. He believes that we are stepping stones leading towards more complexity and we should be happy with our role.

Seeing an idea in the light of another idea can provide new insights. Lee Smolin says in the book What is your Dangerous Idea: “Seeing Einstein in the light of Darwin suggests that natural selection could act not only on living things but on the properties defining various species of elementary particles.”

Seeing the solution to problem with Internet Trolls in the light of a Machine Troll may provide some ideas that can be used as a solution to the problem with Internet trolls.Writing the code that the machine will use is a way to gain insights into human minds. What do you need to include in a code to make sure that the machine acts like a human Internet troll?

Photo:Robots In Bright Colours by Victor Habbick