Online Courses, Schools of Realist Thinking, and Corrigibility in AI
AI Courses you can consider
A few weeks, my supervisor in the firm I work for told me to put together a list of AI courses I would be interested in taking. I’ll be sharing a list of those courses and links to where you can find them. Some of these are organised in learning tracks. So, if you complete a set of courses, you earn a professional certificate or a specialisation. You’ll notice that some courses cut across certain tracks. So, when you do one track, it qualifies for another track.
Let’s get into it:
IBM AI Foundations for Business (Specialisation)
a. Intro to AI
b. What is Data Science?
c. The AI Ladder: a framework for deploying AI in your enterprise
IBM Applied AI (Professional Certificate)
a. Intro to AI
b. Getting started with AI using IBM Watson
c. Building AI-powered chatbots without programming
d. Python for Data Science, AI & development
e. Python Project for AI & application development
f. Building AI applications with Watson APIs
IBM AI Foundations for Everyone (Specialisation)
a. Intro to AI
b. Getting started with AI using IBM Watson
c. Building AI-powered chatbots without programming
Google Digital Transformation using AI/ML with Google Cloud (Specialisation)
a. Business Transformation with Google Cloud
b. Google Cloud product fundamentals
c. Managing machine learning projects with Google Cloud
Lund University
- AI & Law
Learn Quest Ethics in the Age of AI (Specialisation)
a. Artificial Intelligence algorithms, models, and limitations
b. Artificial Intelligence data fairness and bias
c. Artificial Intelligence privacy and convenience
d. Artificial Intelligence ethics in action
You can pay for these courses or apply for financial aid.
Schools of Realist Thinking in AI
If you had any exposure to philosophy while in school, you probably interacted with schools of thought on this or that. In AI, we also have schools of thinking. There are three schools of realist thinking which I’ll introduce in this mail
- Technological Realism
- Societal Realism
- Philosophical Realism
A. Technological Realism
This school believes in the use of open-source frameworks for building AI. Proponents of this school concede that open source comes with the risk of hacking or cyberattacks, but they believe that this weakness is also a strength. The open-source approach ensures that when solutions or upgrades are developed, they can be quickly scaled and adopted across the network of users. Their approach to AI is one of cautious optimism. A chief proponent of this school of thought is John Cohn, computer engineer and chief agitator of the IBM IoT Division, which uses Watson AI
B. Societal Realism
This school sees development as the result of the clash of cultural clashes. A crisis is a precursor to innovation. The anti-authority hip culture of the 1970s led to the microprocessor ecosystem in Silicon Valley, leading to the personal computer. In the same vein, developments in AI will be in response to societal crises. Its chief proponent is John Markoff, tech reporter with the New York Times
C. Philosophical Realism
This school sees fears about AI as baseless and a waste of emotional energy, similar to fears about the Y2K bug at the end of the 20th century. As AI progresses, the belief is that the necessary safeguards will naturally be built alongside, without additional purpose-built rules or developing new philosophies. Our fears about AI stem from our ingrained parochial and patriarchal view of power. The chief proponent of this school is Steven Pinker, bestselling author and cognitive scientist at Harvard University.
Which of these schools do you agree with? I prefer the societal school and partly agree with the philosophical school. I’d love to hear your thoughts on this topic. Feel free to respond to this mail with your opinion.
Corrigibility in AI
What is corrigibility, and what does it have to do with AI?
An AI system is corrigible if it cooperates with its creators when they seek to correct its workings. The opposite of a corrigible AI system resists attempts of its creators to modify or be shut down. The debate of corrigibility presupposes AI systems that are advanced enough to develop a form of consciousness and develop their own goals, distinct from the goals of their creators.
The E-Library
The E-library of AI materials has considerably grown since the last time you may have checked it. Lately, I’ve done a lot of research on AI in healthcare, advanced robotics, autonomous driving and quantum computing, to name a few. As a result, there should be over 300 AI materials for your study and aid your research.
I look forward to hearing about what you’re working on.
You can access the e-library via https://bit.ly/3f0JDjY
