AI: a Tool, not Technology
AI-ML : a Tool, not Technology
Deepam Mishra
AI will be a ‘simple calculator’ for future generations
Not another AI forum!
This is different. This is a platform to focus 2 things
- AI-ML as a tool, not technology…
- With a focus on User-Innovators (not AI-Innovators)
Who This Forum Is For:
For 99% of the future generation. i.e. for the User-Innovators and not the 1% of Maker-Innovators.
Within a decade (or less), AI-ML will become as much as a tool, as Calculators and even PCs are today. I submit that 99% of the next generation will not even know the difference between supervised and unsupervised NNs, AlexNet and GoogleNet etc. However 100% of them will be users of AI-ML.
What will AIML be in this future?
ML-model will be a ‘Calculator for small-x’.
I think that we are still in the toddler days of AIML. It is amazing that today, a computer can recognize dogs from cats, classify squares, circles and stars separately etc. However, these are not very complex tasks for humans. The big-deal, of course, is that these tasks are new for computers. Which means that we can now do these tasks at massive scale with consistent quality. Hence each successful ML-model is fast becoming a ‘Calculator for x’, where x is any task that computers could not do earlier.
Why Should User-Innovators Care?
User-innovators are the ones who make a ‘Big X’ from ‘small x-es’.
History is witness that most maker-innovators focus on a ‘technology x’ and user-innovators in ‘business X”. This is the way the world works, and why would it be different for Calculators for x.
User innovators make Big X’s from small x
The Topics We Should Discuss
For the user-innovators, I would like to wade into the following discussions (other suggestions welcome)
a) Identify Novel Calculators Early. User-innovator discussions on which new calculators are likely to be Big and why. Talk to maker-innovators behind every such new calculator about their intuition and insight that users can leverage.
b) Discover Big X’es: where can each new innovation be used? What are its unique strengths and limitations? Many AIML models are black boxes or at least, grey-boxes and can lead to dangerous outcomes (e.g. racial bias with image classifications etc.).
c) How to Adapt or Re-purpose Calculators: AIML lends itself nicely to simpler (relatively) Transfer Learning or Self-Learning possibilities, that User-Innovators can leverage.
d) Inspire Myself and Future User-Innovators — dispel the fear of AIML. Every innovator is an AI-innovator — whether a gear-head or a poet.
Ergo, The Structure for This Blog
Each article will attempt to follow this structure
a) A thought provoking discussion — e.g.with leading Maker-Innovators and User-Innovators
b) A muscle-building exercise — e.g. specific DIY suggestion for user-innovators (an inspiring video, a Jupyter notebook, a Github etc.)
c) Being vulnerable. Being ambitious. Make bold, albeit wrong, attempts at changing the world. Hypothesize future ‘Big X-es’ within hands-reach or the cliff-edge behind a fog.
