Three features any AI system MUST have
Image ? Sriharsha Ganjam || sriharshaganjam.com

Three features any AI system MUST have

Have you ever wondered how we befriend someone? What are the characteristics that we place the greatest importance to, when presented with a potential new mate? When you approach this question in a logical manner, without getting carried away by the superfluous attributes like wealth, looks, power, etc, we realise there are only 3 attributes that really really matter

  1. A true friend never undermines you and always respects you
  2. A true friend always acts responsibly when dealing with your emotions
  3. A true friend can always be trusted to do the right thing for you

I hope you agree with me on the 3 points above since it's in our human nature to subconsciously scan for these 3 attributes whenever we meet someone new before calling him / her "my best friend".

Off late there is a new entrant in the market seeking our friendship, all we need is for the chemistry between us to click, and this new contender for our affection, can easily become our new Bestest Friend Forever! This potential new friend is neither a "he" nor a "she", but rather an "it" and it is both Artificial and Intelligent. We all call it "AI". We are all impressed by it's power, promise and potential but, and it's a big BUT, we still have our doubts on this new butt less friend (poor joke, I know ??). So does AI have what it takes to be our new BFF?

We are still in the early days of implementing AI across our life, and there are a lot of new opportunities that we can leverage AI for. It's at this stage that we need to be extra careful on how we build a robust AI infrastructure that is inclusive of the above three attributes. We should build systems that respects us to never undermine our decisions, they should never acts irresponsibly towards us, and finally, never compromises on our trust. But, incase you haven't already noticed it yet, these are attributes that cannot be coded and AI is nothing but a massive computer code that is trying to behave like a human. There is no way to programme Respect, Responsibility and Trust into the source code. So, how do we do it?

The solution to this does not lie in the computer code itself, but is rather present outside the code. What we need are thought leaders from us humans to recognise the limitations of AI and draft laws that will uphold our human values and qualities. We need to draw from our collective intelligence regarding earlier innovations and failures to build safeguards to prevent this promising opportunity from derailing. It's similar to how our banks thought us to respect individual privacy by building security firewalls through mobile phone based OTPs that we do not share to do secure online transactions. It's similar to how iTunes and Spotify thought us to consume entertainment more responsibly by paying for subscriptions instead of downloading free pirate copies. It's also similar to how Amazon thought us to trust parting with our money to buy a product that we have never touched or felt before. They were all done through a combination of technological innovations and human governance. We should make the system as foolproof as technically possible but there will always be areas where the technicality won't reach, and for those grey areas, we augment the system by building Human based guardrails. We have all seen what AI systems have done in such a short time, and it is nothing short of amazing. AI has made headway into areas that we previously could not have even imagined just a few years back, but the real magic (and it's a big butt once again! ??), is yet to come. And, it will happen only when we can TRUST our new friend. If not, AI will always be a tool that we only exploit and not a true friend who arguments our efforts.

What Say?

Let me know what you think of in the comments below.

#AI #artificial #intelligence #rules #governance #mind #machine #laws #robotics #ethicalAI

要查看或添加评论,请登录

Sriharsha Ganjam的更多文章

  • The Quantum Dichotomy

    The Quantum Dichotomy

    Given the rate at which things are moving, it may not be too long before we are out shopping for our first quantum…

  • 5 Types Of Disruptions In The AI world

    5 Types Of Disruptions In The AI world

    The only thing constant in this world is Change. The internet is abuzz with all the recent turmoils in the AI world and…

    3 条评论
  • Job losses and AI Agents!

    Job losses and AI Agents!

    Microsoft recently launched the Copilot 365 chat with an Agent Builder that enables users with to build and train AI…

  • Driving an AI Agent

    Driving an AI Agent

    A lot is being said about the upcoming AI Agentic revolution these days, so much so that the year 2025 is touted to be…

  • The Growth Of Agentic AI

    The Growth Of Agentic AI

    There is a lot of excitement on Agentic AI these days and it's barely been a couple of years since AI became…

  • Something that nobody told you about RAG

    Something that nobody told you about RAG

    What is a Question..

    1 条评论
  • Fighting AI Recalcitrance

    Fighting AI Recalcitrance

    Machine Learning is specific subset of AI that allows computers to learn from data and make predictions without being…

  • From Sight to Vision

    From Sight to Vision

    All the complex lifeforms on our planet have 5 sense organs. The power of Sight, Sound, Smell, Taste and Touch.

  • What's, Why's and I's

    What's, Why's and I's

    Growth is universal. From the cosmos to our very own children, everything grows! When Human's first emerged a couple of…

  • From Forecast to Foresight!

    From Forecast to Foresight!

    We all know what "AI" stands for right? And we all think we know why the "A" in AI expands to "Artificial". But what if…

    3 条评论

社区洞察

其他会员也浏览了