Artificial intelligence and the professional

ai

Artificial intelligence is an inappropriate and flawed tool when it comes to decision-making over people issues, such as if child should be taken from their family into protective custody.

It is likely that I and my business will contribute to the advancement of artificial intelligence (AI), so I have a good idea about the limits and potential of this technology.  I wanted to touch on how AI should be a tool or assistant to a professional rather than replacing or undermining that individual.

It’s the great fashion in recent years that everyone gets into AI, which usually means either they have something that automates a system, or it is a pattern recognition tool that pulls conclusions out of big data fed to a network, which acts on the conclusion.  There is a lot of people in government, business and public services who have been sold a bag of poop that AI will save costs and provide a better service if it was used to replace people in the decision-making process in people-related situations.  For example, recruitment by big corporations is now increasingly being automated by AI, so that unless you know how to game the system, it will work against you, and people are reduced to the level of cattle in the corporate system.

There is an obsession with big data, which always has to be cleaned up by low paid humans in places like India to be useable in a pattern recognition system.  These pattern recognition systems such as neural networks operate according to hundreds and thousands of data points, building up through statistics a model upon which conclusions and decisions are made. These models and processes are so complex that not even the designers know how they come to their conclusions, what is called a black box situation.

These models are being used to make life changing decisions about people and their families, for instance if a child should be taken into care, or the appropriate penalty in a criminal conviction, or if someone should be liable for parole.  This impacts me too, I have today been to my first meeting with medical professionals, who consider I should have an autism assessment, but I also shared things like I suffered depression and had thought about suicide.  All I know, this information I shared is being fed into an AI system and it might spit out some conclusion that might lead to me being sectioned by the end of this week, all based on an AI data model rather than human decision-making.

If the reader has coded anything, they will learn that bad code and inputs result in bad outputs.  For example, if I dumped into an AI system voting intentions of a large sample of voters in Clacton UK, and used this to predict how the UK will vote in an overall general election, it might suggest UKIP would form the next government, but when the prediction is tested in real life, UKIP will if they are lucky only have control of the Clacton seat in Parliament. In a rising number of cases it has been discovered that the models built on big data are faulty, biased against certain groups, and are unable to handle unique situations.  People are forced to conform to a narrow set of categories to access services or be on the good side of a statistical artificial computer model that has no relation to reality.

It is a tragedy that for reasons of money, faith in a flawed technology, and a lack of trust of the wisdom and knowledge of human beings with decades of experience in their fields, the AI has replaced the human with tragic consequences for individuals and society.  Families wrongly suffer their children being taken into care, or being imprisoned because the computer judged according to its model this was the right outcome, and nobody can challenge the system data model, because nobody understands how it came to the conclusion.

This is never the way to go for AI, a great tool if used correctly, but totally inappropriate in people-focussed decision-making.  The AI is a useful tool or assistant where the human takes the lead, enhancing their decision-making, for instance in project management, not in decision-making when it comes to people.

Advertisements

Why I support bottom-up technology

internet

Good technological-information systems use the bottom-up approach.

The problem of top-down focus rather than bottom-up in technology, especially in artificial intelligence.

Firstly, this is how I design a system: (example, a system for teaching students.)

Model

This contains the material to be taught, and the strategies used to teach that material.

 

User

The individual student account.

Executive

The brain of the system that processes inputs and outputs between user and the model.

Face

This is the outward face of the system the user sees in which they make their inputs and receives their outputs.

The challenge of top-down approach

Most organisations use the top-down approach.  They make little or no difference between the user and the model.  The user to these people is another dataset in the model that the system tries to force into the model.

Models often use large datasets upon which the system creates stereotypical assumptions based upon the bias of the data.  For instance, in the USA the majority of the prison population is non-white, and the majority of the general population is Christian and white, thus the model is biased for the majority and against the minority.

In the top-down model of a school the bias is against those with learning challenges such as ADHD, Dyslexia or autism.  In the USA and the UK the focus is on results, with an increasingly number of schools run by private companies such as the chartered schools in the USA and the academies in the UK.  All these companies act for profit and focus on results, using a top-down system that is biased against those different from the stereotypical model.  The desire of such schools is to penalise, push aside or get rid of those that are outside of the model.

A common challenge with the top-down method is that the user and the model has been lumped together in the system, and if the user fails to match the model, they suffer in the system.

Using the bottom-up approach

The user is split from the model, the executive focuses on harmonising the model to the unique needs of the user.  If the user has a learning challenge such as ADHD, the executive flags this up to the user account, applies different rules from the model with the appropriate strategies for ADHD students.  The focus is on the user, the model is deployed to the needs of the user.

Conclusion

If someone desires to use my services in designing their system, I will look at their system.  If I see the top-down approach, I won’t work with them.  I think top-down systems belong to a barbarian age, I support the beneficial bottom-up approach to system design.