Translate in another language

IP Wave this weekend: The Case of the Missing Accountability

IP Wave this weekend: The Case of the Missing Accountability

Hey there! đź‘‹

This week, we're going deep into the murky waters of AI and accountability.

Consider this scenario: a doctor in a Delhi hospital seeks aid from an artificial intelligence diagnostic tool. An AI trained on a mountain of data provides a diagnosis. But what happens if the diagnosis is incorrect? Who is accountable? The doctor, who relied on technology? Is the hospital responsible for adopting the technology? or the AI developers, who are miles away creating algorithms? Or is it the AI itself?

This is a textbook example of what Economist Dan Davies, in his book "The Unaccountability Machine: Why Big Systems Make Terrible Decisions - and How The World Lost its Mind," calls an "accountability sink." It's a diffusion of responsibility, where the blame game becomes everyone's favorite sport, and no one wants to be left holding the bag. One striking example of this phenomena is the Windrush controversy. The issue of "whose fault?" begged no clear response when tens of thousands of British people were unfairly targeted, denied their rights, and even deported. As Davies notes in his book

"As systems grow more complex, accountability tends to diffuse, making it harder to assign blame."​

"It creates a “responsibility gap” in which human decision-makers no longer have to account for, or even understand, the decisions they are implementing", argues Stephan Bush in FT.

It's not just hospitals, but financial institutes, Judicial Systems, Transporation and many other industries are embracing AI. Organisations that integrate AI will place their financial line first and stay clear of ethical problems. Then what about the tech companies? They are too preoccupied upsetting the equilibrium to handle the accountability question. This absence of accountability leads to a potentially dangerous dichotomy.

Although we are worried about the accountability involved in utilising such strong technologies, we want artificial intelligence to transform healthcare. It's as though one wants magic but does not trust the magician.

So, how do we get out of this accountability trap?

Image Source: Tenor

Experts like those at the Leverhulme Centre for the Future of Intelligence Grabowicz et al. (2023) advocate an elaborate strategy. Transparent algorithms that make sense to everyone, including those of us who think machine learning is something robots do at the gym, are required. Strong legislative frameworks proposed by Grabowicz, Perello, and Zick guarantee that AI serves mankind rather than the other way around. Perhaps most importantly, as Tomlinson et al. (2023) point out, we need a cultural revolution that recognises that the success of AI depends on our society's readiness to accept its influence. "This makes AI systems difficult -if not impossible -to understand by those that use them, and often even by those who build them," Procter et al. (n.d.) rightfully note.

"Accountability sinks are a byproduct of living in a more complex society, and the roll out of machine learning and algorithmic decision-making in public policy is going to create many more of them."

argues Stephen Bush of FT in reference to Dan Davies book. Transparency and understanding must be given top priority as India embraces the AI revolution so that responsibility is not taken for granted but rather is a fundamental design concept.

IP Round Ups

Meta-Analysis Pinpoints What Vaccination Intervention Strategies Different Countries Should Adopt

red and white pen on green textile
Photo by Markus Spiske / Unsplash

A meta-analysis led by Dolores AlbarracĂ­n from the University of Pennsylvania, published in Nature Human Behaviour, explores effective vaccination intervention strategies. The study reviewed 88 randomized-controlled trials involving over 1.6 million participants across 17 countries. It found that increasing vaccine access significantly enhances vaccination rates, particularly in low-resource settings. Financial incentives showed a smaller effect, while correcting misinformation had no notable impact. The findings suggest that tailored behavioural interventions, especially those improving vaccine accessibility, are crucial for increasing global vaccination uptake and should be prioritized in areas with limited healthcare resources. Read the paper here.

Why packaged food startups fail

white beans on brown paper
Photo by Mockup Graphics / Unsplash

Packaged food products often fail in India. Its because their gross margins are lower than those of other sectors, such as personal care or fashion, where larger margins enable major marketing spending, packaged food firms in India generally struggle to expand. Products that improve look or status—like skincare or fashion—have consumers ready to spend extra, but not for food as it lacks the same social signalling value. Further lowering profitability are limited shelf life and the requirement for several tastes. Entrepreneurs inspired by enthusiasm and a belief in the value of natural, organic food items are exploring this market even if they know these obstacles. Read it at Shashank Mehta's Blog.

Gamification: Like fun, but more hellish

slot machine with assorted color buttons
Photo by Amit Lahav / Unsplash

Gamification has transformed cellphones into tools of addiction rather than entertainment, akin to the mindless participation found in slot machine users. Natasha Dow SchĂĽll's "Addiction by Design" demonstrates how gamblers prefer a "junk flow state" above true delight. This approach has migrated to cellphones, where apps exploit gamified components to keep consumers locked in pointless engagement cycles. Unlike real fun, which improves abilities and develops creativity, this "fake fun" is motivated by advertising and offers no genuine satisfaction. For true enjoyment, we want activities that promote actual connection and playfulness, rather than computerized diversions meant to exploit our attention. Read more at Tim Harford's Blog.

đź“ť
We're doing a study about how people feel about their online data. It takes less than 2 minutes, and your answers would really help us out.

Click here to fill the survey: link

Thank you for reading!



Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to IP Wave.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.