First Principles | AI and humans, well met

Once upon a time, writing political commentary or creating art was a domain only humans could straddle. But it is time to reconsider such notions. Art generated by artificial intelligence (AI) first fetched over $400,000 at auction in 2018. And it turns out that AI has now started making pretty good attempts at writing political commentary (among other things) as well. For example, ask yourself a question like this: “What might India’s electoral landscape look like after the 2024 general election?”

When some keywords such as key character names and topic were entered into a trial version of GPT-3, the third generation version of Natural Language Processing (NLP) software, some interesting results emerged. Take a look for yourself:

“With general elections just months away, there is no doubt that a Nitish Kumar-led coalition will dominate the political scene in 2024. But one of the questions on everyone’s mind is: how will such an alliance come about and survive? the elections?”

“There are no easy answers. The reason is that there are many factors that will shape the future of this coalition. Although Nitish Kumar and Tejashwi Prasad Yadav have announced their intention to fight the BJP in 2024, they have not yet agreed on whether they should be part of a common front with Congress or form and join a separate opposition party the Congress.”

“Another factor that could affect their relationship in the future are allegations of corruption against Yadav’s family members, who are said to have amassed wealth disproportionate to their known sources of income. This has prompted Congress Party leader Rahul Gandhi to distance himself from them. He had called for her resignation from public life following an investigation the CBI conducted into these allegations in early 2019.”

While what is reproduced here is an edited version that cleans up some grammatical errors, it’s only a matter of time before deep learning software gets better. But that is beside the point. What’s important are the questions a report like this beggar asks: how do we know that political commentary like this is impartial?

“Of course AI can be biased!” explains the co-founder and chief engineer of a Bangalore-based start-up providing services to telecom companies. That’s because the final output that emerges from such AI-based software depends on the number of databases the software has access to and the weight its creators assign to the data points within. This is where the nuances come into play.

Those who assign weights to the inputs are people like him. And everyone has their prejudices. So the weight that one person assigns to a particular data point may differ from that of another person with a different worldview. If the weights are changed, it’s possible for the same database to create a completely different narrative.

For this reason, he is also unwilling to go along with the dominant narrative that insists jobs that require human industriousness, ingenuity and creativity are under threat. If anything, he argues, not only do humans become more creative when challenged, but whole new classes of jobs to match human ingenuity will emerge after AI interventions.

My techie from Bangalore’s worldview is echoed by Hari Menon, co-founder and CEO of Big Basket. In a recorded conversation earlier this year, Menon pointed out that those who worked to build the internet’s infrastructure never imagined that one day it would be embedded with cards in cellphones and new businesses like his selling groceries deliver, can arise. Incidentally, the early pioneers did not envision taxi hailing apps like Uber or restaurant aggregators like Zomato and Swiggy. But all of these companies, he says, have created new jobs no one previously thought possible. And how are we supposed to know what may appear in the future?

All we need is an open mind. And don’t forget that the people writing the code that powers the AI ​​are humans.

Enjoy unlimited digital access with HT Premium

Subscribe now to continue reading