The AI Progression (ANI, AGI, ASI)

When Gen-AI & its progression is largely targeted at improving productivity this is how it plays a role in affecting the expectations of the users…

The Narrative:

In the recent past when I was conversing with someone in his late 50s, I learnt how he was totally amazed over his kids (in their early teens) raving about ChatGPT & how it has made their life simpler. He mentioned how their whole work was completed within a few clicks & although the number of hours they spent over their phones had increased 2x now. When the quinquagenarian surely seemed impressed by the advancement in technology & how it has penetrated our lives, he also expressed a sort of a fear over how kids today don’t seem to spare a thought & focus on the learning part, are getting sucked in & becoming over-reliant on AI.

Of course, overreliance on AI seems to be a BIG problem across age-groups today. Many people trusting it blindly seems to be the scariest part here. But relevance & accuracy when handling ridiculously simple & unbelievably complex queries seem to be major deterrents as mentioned over this article published by Stanford vide a case study of a Maze of varied complexity.

The number of AI tools released have seen an alarming rise & are still northbound as we speak. Another thing that’s also seen an astronomical rise alongside these tools is the expectations of the markets and the users per se. There’s no doubt on how these tools have made so many users’ lives simple & they seem to conquer newer areas by the day.

• So, given how the lives of people have got simpler today with AI how should product leaders & XFN teams factor all that into their workflow while building products?

• Would it be right to say that building a product without employing AI could eventually witness meagre adoption rates post release & may not augur well for the org. building it?

I’ve personally been a part of many discussions, arguments, polls that try to find answers to the question:

• “Is AI going replace product managers?”

or a more common one that is:

• “Is AI going to take my job away?”

And the most popular answer was on the lines of:

• “AI may take your job / replace you unscrupulously if you put a deliberate effort to keep away from it & find some lame reason to not adopt it into your regular workflow turning you unproductive in comparison to the others”

Take for instance, writing business documents like say the PRD. Post gathering all the inputs & having enough info to work with, an average a product manager may take about an hour to build one & about 2-3 hours in the worst case. But with the right prompt the very same PM could perhaps generate a better PRD within a matter of seconds using AI, which stands proven here as for this graph published by the Nielsen Norman Group.

Although most of what we see around us today falls under ANI (Artificial NARROW Intelligence) the one thing that PMs & leaders ought to consider is the coverage & inclusion of AI reaching the grassroots intern bumping the expectations of the users by 3x. The users’ expectations have undeniably risen astronomically in the last 3-4 years & what could have resulted in good sales revenues with a mere satisfactory product about a decade ago could get turned down / rejected outright as dissatisfactory in today’s time.

Definitions:

ANI (Artificial NARROW Intelligence)

This is the AI that is ubiquitously found around us today. It doesn't boast of complex / unbelievable capabilities & could carry something of narrowed down focus to a specific task. ANI is also referred to as Narrow AI / Weak AI. Amazon’s Alexa, Apple’s Siri could serve as examples of ANI.

AGI (Artificial GENERAL Intelligence)

This could refer to AI that comes induced with a load of cognitive capabilities, which could largely challenge the capacity of human intelligence. Most of this Tech is still in the conceptual stage & it doesn't have a clearly defined scope / boundary. Cognitive Vision based products, Autonomous / driverless vehicles could fit the description of AGI.

ASI (Artificial SUPER Intelligence)

This is the AI that could fit an artistic description of what the Tech could be capable of & a figment of one’s imagination. It is not only believed to match but go well beyond the brightest (not average) human intelligence that’s known to man today & the scope could be much wider than just a specific task / problem that AI around us is known to address today. The robots featured in popular Sci-Fi movies like the Terminator series could serve as befitting example here.

When there are enough posts talking about the AI tools that one MUST put to use across the PLC so as to target and improve productivity, over this article I choose to focus on alternative possibilities / improved scope.

UXR (UX Research)

No doubt, AI could act as a superlative tool over conducting user research given how UXR itself seems to have moved from the conventional methods that involved hitting the users up in-person 1:1 with questions & then being able to dig in deeper following them up with another set so as to build a granular / personalized understanding.

But with the advent of AI some bits of UXR like sentiment analysis & usability testing are easily automated with a significant improvement in the TAT (turnaround time – time taken for research to go from 0-1) & the quality of insight gained.

Some areas where AI could help UX Researchers are:

Design

AI is known to save designers a lot of time as well given the ample choices on offer over the whole design process although the creativity part is still largely is the designer’s turf.

AI arms designers with tools that could help them generate images / visuals / videos in a click by way of proper prompts & also helps them focus more on the solutioning bit by abolishing the worries about other peripheral stuff like say, alignment over a given frame / webpage / mobile app.

Development

The time required to build a SPA (Single Page Application) has been going drastically down owing to how there are plugins / frameworks / boilerplates made available across a wide array of platforms strongly discouraging the need to build everything from scratch.

Taking the reusability feature a notch higher from what was merely code reuse (Git / GitHub etc.), AI offers developers with tons of other options that’s to do with analyzing, debugging & maintaining code efficiently.

---

The Verdict:

The scope & coverage of AI is increasing as we speak. When the limitations over its scope could get blurry over a period of time, the wisest & foremost thing to do it is to understand where it would work and where it wouldn’t.

When staying aloof is not an option anymore, there’s still a visible gap between the actual capability of the Tech & the current usage patterns. And of course one thing that ought to go without saying is, use it, but don’t use it blindly.

---