Viswanathan K's Thoughts

During Donald Trump’s tenure, there was a big fight between the US and China about Huawei , the Chinese telecom equipment (and smartphone) manufacturer. The West was worried that if Huawei equipment grabbed the largest share of the upcoming 5G market, then China could insert malware and spyware in telecom networks all over the world…   The West would know. As Anirudh Suri writes in The Great Tech Game , in the 19 th and 20 th centuries, Britain had the monopoly over “telegraph communication network” across the world. Progressively, physical infrastructure monopoly had resulted in a monopoly over the raw materials needed for telegraph cables. As Britain became the #1 player in telegraph systems, it was cheaper and more economical for other countries to use British systems. With ever larger systems under their control, British expertise at laying cables and repairing these systems became better. It had become a circle that reinforced British dominance over telegraph systems across

Imagine a teacher who asks a maths question to the class, then gently nudges the students towards the correct answer (affirmation when they are on the right track; correcting them when they seem to be going off the path). A variant of the above approach is what David Epstein describes in Range : “’Lemme show you, there’s a better, easier way.” If the teacher didn’t already turn the work into using procedures practice, well-meaning parents will.”   Next imagine a second teacher who lets the students try and solve it on their own. No feedback in real time. Afterwards, she corrects the paper and includes notes on the right approach (if needed).   Which way do you think results in better learning? Not just for the duration of the class or course, but in the long run? It’s the let-them-struggle, let-them-fail approach that yields better learning, not just in that course/class, but in long term retention as well!   There’s even a term for it, the “hypercorrection effect”: “Th

Hollywood is full of apocalyptic movies where the machines and AI takes over the world – the Terminator series is one of the best-known examples. In recent times, as machine learning algorithms get better at almost everything which seemed impossible just a few years back – recognizing faces, photos, transcribing and even translating spoken words in real-time – many people have been voicing their concerns, from Elon Musk to Bill Gates. There’s a term for it – the “singularity”. It’s the point of no return, at which technology takes over… for good.   On the other hand, China, which is second only to the US in all matters AI, doesn’t seem worried. In fact, tech entrepreneurs in China are optimistic that AI advances will make life better. Why the difference, asks Kai-Fu Lee in his book . For one, it’s their experience with technology so far: “The Chinese government has long emphasized technological advances as key to China’s economic development… For the last forty years, Chinese pe

Earlier, I spoke of the importance of mental models and feedback in designing user friendly products. Let’s take a few examples Cliff Kuang’s User Friendly .   He cites driverless cars as an example of the challenges when something is new. And how Audi went about handling it. First , it should be obvious when the car is in auto mode (Mode confusion has led to many airline crashes). Audi needs two buttons need to be pressed together to transition to auto mode – this prevents accidental activation. When the car takes over, the color of various panels changes to convey the status. Second , the occupant should know what the car is going to do, before it does it. Surprises don’t sit well with user experience. Hence, before changing lanes, the Audi shows a countdown timer informing what it is going to do next. Third , one should be able to “see” what the machine is “seeing”. Else, one is nervous what the car may be missing. On the display, the Audi shows all the cars around it. Fourth ,

I bought Tim Marshall’s account of the Yugoslavia war in the 90’s, Shadowplay , because I don’t understand anything about the place. Or as Marshall put it: “I thought I knew my history, but actually coming to a region where everyone seemed to have a grievance and an ‘itch’ at the end of their name was confusing. Milošević. Panić, Ilić?” In case you’re wondering, this is not a popular history book. Instead, it’s a British journalist’s account of his stint in Yugoslavia during that period.   With typical British wry humour, he pointed out Europe’s surprise by the carnage that broke out after the death of Marshall Tito who had convinced folks that “they really were Yugoslav first, Croat/ Bosnian/ Muslim/ Serb second”. But after his death and the fall of communism, things old divisions resurfaced: “To my generation it just didn’t seem possible. War was what happened far away, in places with different cultures. War did not happen in our continent because we’d left all that behind

Designing for the user is easier said than done. That’s obvious. In User Friendly , Cliff Kuang points out a key element to such design: “(It is key to understand) the ways in which humans assume their environment should work, how they learn about it, how they make sense of it.” If the user can’t make a mental map or model of the product, he’ll struggle to use it. Put differently, the device should work the way the user expects it to work. If you don’t see the importance of the mental model, consider the Internet.   Those of us who are started using the Internet when it started to take off think of it as a set of sites with links between them. The browser was a way to navigate across sites. That’s our mental model of the Internet – as the World Wide Web (sites linked to one another).   The majority in poorer countries, though, did not get to use the Internet until the smartphone became ubiquitous. Their model of the Internet is nothing like a “web”. To most Indians, WhatsA

Why is it that India is the only country with a government created smartphone payment system (aka UPI upon which all mobile payment apps from from PayTM to PhonePe to BHIM to Google Pay are built)? How come China’s private sector payment apps (WeChat and AliPay) are now so ubiquitous that cash is not even accepted in more and more places across the country? The other side of the coin – and the very curious one – is that no Western country has built any such smartphone based digital payment system. What’s going on?   I thought the answer lay in the fact that MasterCard, VISA, and the banks of the West lobby against any such system since it would eat into their commissions. While that’s definitely part of the reason, Anirudh Suri’s book, The Great Tech Game , reminded me that there are other reasons as well.   An additional reason why countries like India and China took the move, writes Suri, is that the current banking system to move money across countries, the SWIFT system, is

The secret project was called “Olympic Games”. Its aim was to cripple the Iranian nuclear program “without setting off a regional war”, writes David Sanger in his super-interesting book on cyber warfare titled Perfect Weapon . The US and Israel settled on creating a malware (computer virus) that would speed up/slow down the Iranian nuclear centrifuges, leading them to “ultimately destroy themselves”.   Being a covert operation, the US (and Israel) couldn’t claim they’d done it. So how then did this story, the malware now known as Stuxnet, break out? Well, it was rooted in the fact that Stuxnet couldn’t be simply added onto an Iranian centrifuge (obviously). It had to be spread all over the world, and hopefully would end up entering the centrifuges via malice (an Iranian traitor) or by usual stupidity (someone carrying an infected USB into work). As with any scatter and pray operation, the malware thus reached all over the world (We’ll come to why it didn’t do any damage anywhere el

Cliff Kuang’s book, User Friendly , is an interesting romp of how much the design of products has evolved to make them easy to understand and use. It wasn’t always that way. Once upon a time: “(The view was) that correctly operating a machine was about finding the right person to operate it.” ‘Pilot error’ is the term that sums up that attitude – if something went wrong, it was the user’s fault.   The World Wars began to change that. Why? “The performance of men under stress bore no resemblance to that of those operating a demonstration model.” And from that emerged the idea that machines could/should be designed to “better conform to the limits of (humans’) senses and minds”.   That mindset eventually spilled over into consumer products. The driver was consumerism – businesses could see that user friendliness could be the “elixir of sales growth”.   But it needed the “profusion of computers and electrical gadgets” for the trend of designing for the user to really take off. And now, th

In 1994, around 800,000 people were killed in the Rwandan genocide. In response, the UN decided that if a country couldn’t provide physical and economic security to its citizens, other countries (with some checks and balances) could intervene to restore order, writes Richard Haass in his book, The World .   This came to be called the Responsibility to Protect (R2P) doctrine. Since it was such a grey area (how much disorder justifies external intervention?), it was never put into practice. Until the US and its NATO allies invoked R2P in Libya in 2011. The rest of the world soon came to see it as just a thinly veiled attempt to overthrow the government of Gaddafi. Even worse, it led to anarchy, the very thing they claimed they were trying to prevent/fix.   With the Libya episode, R2P is dead. Russia and China were now set against R2P, viewing it as a “cover for imposing political outcomes”. The West too realized that no outcome could be guaranteed; it could make a bad situation w

In an earlier blog , we saw how the Chinese government was instrumental in kick-starting China’s foray into AI. But that started in 2014, so how has China made so much progress in such a short time? Given that most of the R&D around AI was done in the West – the US, UK, and Canada - how did China emerge in this #2 position so quickly?   It helped greatly that most of the AI algorithms are public knowledge. They are not trade secrets or protected by patents. Combine that with the fact that these AI’s get better the more data they have access to. Not surprisingly, China with its huge population, generates enormous data.   Further, unlike the West, China doesn’t care about privacy. No, this isn’t just because the government says so. Rather, most Chinese don’t mind sharing their data with Alibaba, TikTok, Baidu or WeChat either. They find the conveniences and features they get in return to be worth it.   Plus, China has learnt AI by doing, i.e., by its entrepreneurs trying

Andrew Sullivan is a right-of-center guy in American politics. His article on the state of American politics makes one realize the left-right polarization is pretty much the same across the world.   Even though he is right of center, he was among those who voted for Biden. Why? “There was no choice in 2020, given Trump. I understand that.” Now he is thoroughly disillusioned with Biden and his policies. He feels that Biden has gone overboard in championing the “entire far-left agenda”. In American politics, being “far-left” means (1) Excessive increase in government spending, which if overdone then leads to inflation. As has happened now; (2) Increasing receptiveness to immigrants; (3) Excessive catering to the gay/lesbian side, far beyond just ensuring that those groups aren’t discriminated against; (4) A belief that it’s better to go soft on crime if the alternative will result in even more blacks landing in prison; (5) Individual debts (e.g. student loans) should be wiped off

Capitalism, in combination of industrialism, led to more and more inequality. Eventually, as Anirudh Suri reminds us in The Great Tech Game , that inequality led to a backlash and to the rise of Marxism and communism. And then communism lost favour, with the fall of the USSR.   Today, we seem to be on the cusp of history repeating itself. Well, not repeating. Rather, as Mark Twain famously said: “History never repeats itself, but it does often rhyme.”   Suri elaborates. The tech companies of today have become unimaginably rich, with the likes of Apple, Amazon, Microsoft, Google and Facebook becoming trillion-dollar companies (not a typo, that’s “trillion” with a “t”). At the same time, these companies don’t generate employment on the same scale as manufacturing behemoths of the past. They do more with far, far fewer employees. Which is why Suri worries: “As technology and capitalism combine to create some of the same structural inequalities and class conflicts, new socio-p

Decades of attempts at Artificial Intelligence (AI) failed, until the advent of “neural networks”. That is basically an attempt to make computers “learn” the way our brains learn. For example, one feeds the system photos along with labels (cat, dog etc) and lets the computer find the patterns corresponding to each. Once ready, one then feeds it new photos that weren’t part of its training. This is the test – to see whether the patterns it noticed were correct or not.   That approach has yielded the AI we see today all around us – speech recognition (Alexa), image recognition (face unlock on your phone), driverless vehicles etc. While that’s very impressive, it is what is called “narrow AI” – limited to specific topics only. General purpose AI is still far away.   The Chinese venture capitalist and ex-Googler, ex-Microsoft, Kai-Fu Lee wrote an excellent book o n the state of AI in China. While the US is Number 1, he says China is the clear Number 2 and closing in fast, even ahe