What is the worst that go wrong with AI and why and how?

The Quora's folks are asking very hard questions and one of them i'd like to share with the EU AI community:

It is already going wrong, for 5 reasons:

There is no general understanding what is real AI;

Real AI is imitated by its smallest field, ML with its special branch of Deep Learning;

No industrial enterprises or commercial entities are doing actual AGI;

Most resources go to military AI;

No AI standards and norms, strategies, plans and policies, on national or international levels.

There is no general understanding what is real AI

Real AI is about building computer systems capable for self-recursive improvement, autonomous updating, upgrading and conscious actions, as a result of their encoded world schema and experiences, as well as discovering the basic rules and smart algorithms that govern its complex intelligence/data processes.

And it is hardly just about” building computer systems that automatically improve themselves as a result of their experiences, and what are the basic rules that govern learning processes”? What is hardly pushed by the machine perception/learning folks.

A real machine learning, as deep neural nets (DNNs), is to create systems that can autonomously learn to perceive the world, on their own. And smart machine implementations — including cognitive robots, autonomous vehicles, virtual personal assistants (VPAs), smart advisors, etc. — act in an autonomous manner in the real world settings.

To be fair, things went wrong from the very beginning, when John McCarthy came with his circular definition while trying to answer Q. “What is artificial intelligence?

A. “It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable”. Basic Questions

Real AI is imitated by its smallest field, ML with its special branch of Deep Learning

Today we witness a big commercial hype by big Internet companies, from Google, Amazon and Facebook to the Chinese BAT, where ML/DNNs promoted as AI, what might end up with a big disillusion, as happens with any hype.

No industrial enterprises or commercial entities are doing actual AGI

A fair picture of the current level of AI maturity in the market is closed and all hype is revolving around some big IT companies currently investing in ML-based technologies.

Many are using AI just as a marketing tool or a buzz word, having no idea of its real nature.

Here is a typical business trick to become an AI firm. One need just “self-identified businesses that claim to be using or building AI”, as suggested in "The New Artificial Intelligence Market - O'Reilly Media".

Its AI Definition: “Technologies available that are achieving human- or super-human-level intelligence on a given human task (aka “narrow intelligence”)”; however, “it is not the purpose of this report to argue what Artificial Intelligence is or is not’. Rather, I take a practical approach to the definition of AI and present an analysis based on self-identified businesses that claim to be using or building AI.”

There are only 1,500 companies in North America that are doing anything related to AI today, even using its narrow, task-based definition. That means less than one percent of all medium-to-large companies across all industries are adopting AI.

And what is going on with all sorts of AI fora is nothing but nonsense, a strange group of people is always presented as world leading AI experts, just to attract investors, Wall Street, venture capitalists, or technology executives. Rise of AI Conference 2019

As smart people joke: the most profitable use of AI - to make Conference.

As a result, people are badly confused.

It is nonsense, but data analytics, software development, IoT companies, etc., are to make AI industries.

If you check “self-designated databases of AI startups” getting round funding, as “Global Artificial Intelligence Database with 3,466 AI companies”, one hardly finds any real AI company.

As a matter of fact, we need first to decide some common criteria for AI and its Industry.

Valuing the Artificial Intelligence Market, Graphs and Predictions | Emerj - Artificial Intelligence Research and Insight

At least, the AI Industry should be upgraded as including:

Augmented Analytics;

Autonomous robots (including self-driving vehicles)

Intelligent Digital assistants (e.g., better Siri, automated online assistants)

Smart Neurocomputers;

Smart Embedded systems (machine monitoring and control systems)

Expert systems (e.g., medical decision support systems, smart grid)

Cognitive robots, (a potential $2 trillion dollar market for “cognitive computing”), etc.

AI arms race

On the top of this all, we have a fierce competition between big powers and other states to have its military forces equipped with the best AI. And a global arms race for best AI has already begun. As the power of narrow AI grows, the likelihood of a future war filled with killer robots grows as well and lethal autonomous weapon systems (LAWs) are going out of any international control. 2018 Group of Governmental Experts on Lethal Autonomous Weapons Systems (LAWS)

Artificial intelligence arms race - Wikipedia

AI standards and norms, strategies, plans and policies, on national or international levels

Only few powers managed to reveal a sort national AI strategy and development plans, as China and USA:

THE NATIONAL ARTIFICIAL INTELLIGENCE RESEARCH AND DEVELOPMENT STRATEGIC PLAN by The National Science and Technology Council (NSTC); USA, 2016

SUMMARY OF THE 2018 WHITE HOUSE SUMMIT ON ARTIFICIAL INTELLIGENCE FOR AMERICAN INDUSTRY by THE WHITE HOUSE OFFICE OF SCIENCE AND TECHNOLOGY POLICY

A Next Generation Artificial Intelligence Development Plan; 2017, by China's State Council

Coordinated Plan for Artificial Intelligence, EU, 2018

So, things already have being going wrong with AI, and the only question what it will cost humanity and how to correct all this commercial hype.

ARTIFICIAL INTELLIGENCE IS THE MOST POWERFUL TECHNOLOGY OF THE 21ST CENTURY!

To play an unfair game with AI is to play with the humanity’s future.

What is the worst that can go wrong with AI and how?

https://www.quora.com/What-is-the-worst-that-can-go-wrong-with-AI-and-h…

https://www.linkedin.com/pulse/artificial-global-intelligence-convergin…

Tagy
EU AI

Komentáře

Profile picture for user isaundda
Vytvořil uživatel Darek SAUNDERS dne Pá, 21/12/2018 - 13:21

I am most concern about reasons 1 and 5: There is no general understanding what is real AI and no AI standards and norms, strategies, plans and policies, on national or international levels. In times of post truth and fake news more needs to be done to educate our communities.

The fact that a large amount of resources goes to military AI is a reality and is similarly happening with many other disrupting technologies. There is not much we can do and additionally we should look at the bright side. Our law enforcement and militaries have to be AI ready otherwise they could be easily overcome by AI supported extremists and rogue states.