What went so wrong with covid in India? Everything.
But voices like his were drowned out by the federal government’s messaging, which suggested that India had somehow outwitted the virus. The hype was so strong that even some medical professionals bought into it. A Harvard Medical School professor told the financial daily Mint that “the pandemic has behaved in a very unique way in India.”
“The real harm in undercounting is that people will take the pandemic lightly,” says Arun. “If supposedly few people are dying due to covid, the public will think it doesn’t kill, and they won’t change their behavior.” In fact, by mid-December India had reached yet another somber milestone: it recorded its 10 millionth infection. It was only the second country to do so, after the US.
The government hadn’t used the first lockdown wisely, but December was its chance to set things right, says Gagandeep Kang, a professor of microbiology at the Christian Medical College in Vellore, Tamil Nadu. She says that a number of tactics—ramping up sequencing, studying public behavior, collecting more data, refusing permission for superspreader events, and starting the vaccine rollout earlier than planned—would have saved many lives during the now-inevitable second wave.
Instead, she says, the government continued its “top-down approach,” in which bureaucrats rather than scientists and health-care professionals were making decisions.
“We live in a very unequal society,” she says. “So we need to engage people and build partnerships at a granular level if we are to effectively deliver information and resources.”
In December the government of Goa let its guard down entirely. The state is heavily reliant on tourism, which makes up nearly 17% of its income. The bulk of the tourists show up in December to celebrate Christmas and New Year on sandy beaches with raves and fireworks.
Vivek Menezes, a Goan journalist, says that the state’s reputation as “the place to be” had not faded during the pandemic. “It’s the place for India’s wealthy and for Bollywood, and therefore it’s the place for India,” Menezes says. The pandemic had kept foreign tourists from visiting, but domestic holidaymakers poured in. Some states, such as Maharashtra, had placed restrictions at their borders; others, like Kerala, had a strict policy of contact tracing. In Goa, visitors didn’t even have to show a negative covid test. And the state’s masking policy extended only to health-care workers, visitors to health-care facilities, and people showing symptoms. “Goa was left to the dogs,” says Menezes.
The world’s largest superspreader
India started 2021 having registered nearly 150,000 deaths. Only then, in January, did the government place its first vaccine order, and it was for a shockingly low amount—just 11 million doses of Covishield, the Indian version of the AstraZeneca vaccine. It also ordered 5.5 million doses of Covaxin, a locally developed vaccine that has yet to publish efficacy data. Those orders fell far short of what the country actually needed. Subhash Salunke, a senior advisor to the independent Public Health Foundation of India, estimates that 1.4 billion doses would have been required to fully vaccinate all eligible adults.
On January 28, in an address to the World Economic Forum in Davos, Modi declared that India had “saved humanity from a big disaster by containing corona effectively.” His government then gave the go-ahead for the Kumbh Mela, a Hindu festival that attracts crushing throngs of millions of people to the holy city of Haridwar in the northern state of Uttarakhand, which is famous for its temples and pilgrimage sites. When the state’s former chief minister suggested that the festival should be “symbolic” this year given the circumstances, he was fired.
A senior politician in the prime minister’s Bharatiya Janata Party told the Indian magazine The Caravan that the federal government had its eye on the forthcoming state elections and didn’t want to lose the support of religious leaders. As it turned out, the Kumbh wasn’t just any superspreader event—with a reported 9.1 million people in attendance, it was the world’s largest superspreader event. “Any person with a basic textbook on public health would have told you this was not the time,” says Kang.
In February Salunke, the public health expert, was working in an agrarian district in the western state of Maharashtra when he noticed that the virus was transmitting “much faster” than before. It was affecting entire families.
“I felt we were dealing with an agent that had changed or appeared to have changed,” he says. “I started to investigate.” Salunke, it now turns out, had found one mutation of a variant that had been detected in India the previous October. He suspected that the variant, now known as delta, was about to run rampant. It did. It is now in more than 90 countries.
“I went to all those who are responsible and those who matter—whether district level officials or bureaucrats at the central level, you name it. Everyone who I knew I immediately shared this information with,” he says.
Salunke’s discovery doesn’t appear to have affected the official response. Even as the second wave was accelerating and after the WHO designated the new mutation “a variant of interest” on April 4, Modi kept up his hectic schedule ahead of state elections in West Bengal, personally appearing at numerous public rallies.
At one point he gloated about the size of the crowd he had attracted: “In all directions I see huge crowds of people … I have never seen such crowds at a rally.”
“The rallies were a direct message from the leadership that the virus was gone,” says Laxminarayan of the Center for Disease Dynamics, Economics & Policy.
The second wave filled hospitals, which quickly ran out of beds, oxygen, and medication, forcing gasping patients to wait—and then die—in homes, in parking lots, and on sidewalks. Crematoriums had to build makeshift pyres to keep up with the demand, and there were reports that the outpouring of ash drifted so far it stained clothes a kilometer away. Many poor people couldn’t even afford to pay for funeral rites and immersed the bodies of their loved ones directly into the River Ganges, which led hundreds of corpses to wash up on the banks in several states. Alongside these apocalyptic scenes came the news that deadly fungal infections were overwhelming covid patients, likely as a result of lower infection control and overreliance on steroids in treating the virus.
Chaos continues; Delta spreads
And all along, there has been Modi. The prime minister had been the face of India’s fight against the pandemic—literally: his headshot appears prominently on the certificate given to people who get their vaccine. But after the second wave, his premature triumphalism was mocked and his lack of preparedness derided widely. Since then, he has gone largely missing from the public eye, leaving it to colleagues to place the blame elsewhere, most notably—and inaccurately—on the government’s political opposition. As a result, Indians have been left to face the biggest national crisis of their lifetime on their own.
This abandonment has created a sense of camaraderie among some groups of Indians, with many using social media and WhatsApp to help each other out by sharing information about hospital beds and oxygen cylinders. They have also organized on the ground, distributing meals to those in need.
But the leadership vacuum has also produced a huge market for profiteers and scammers at the highest levels. In May, opposition politicians accused a leader of the ruling BJP party, Tejaswi Surya, of taking part in a vaccine commission scam. And the health minister of Goa, Vishwajit Rane, was forced to deny claims that he played a part in a scam involving the purchase of ventilators. Even the prime minister’s signature covid relief fund, PM Cares, came under fire after it spent Rs 2,250 crore (over $300 million) on 60,000 ventilators that doctors later complained were faulty and “too risky to use.” The fund, which attracted at least $423 million in donations, has also raised concerns about corruption and lack of transparency.
A successful vaccination agenda might have helped erase the memory of the string of missteps, but under Modi it has only been one technocratic mistake after another. At the end of May, with far fewer vaccines in hand than it needs, the government announced plans to start mixing doses of different vaccine types. And at the height of the second wave, it introduced Co-WIN, an online booking system that was mandatory for anyone under 45 who was trying to get vaccinated. The system, which had been under scrutiny for months, was disastrous: not only did it automatically exclude those who do not use computers and smartphones, but it was also hit by bugs and overwhelmed by people desperate to get protection.
The Download: how we can limit global warming, and GPT-4’s early adopters
Time is running short to limit global warming to 1.5°C (2.7 °F) above preindustrial levels, but there are feasible and effective solutions on the table, according to a new UN climate report.
Despite decades of warnings from scientists, global greenhouse-gas emissions are still climbing, hitting a record high in 2022. If humanity wants to limit the worst effects of climate change, annual greenhouse-gas emissions will need to be cut by nearly half between now and 2030, according to the report.
That will be complicated and expensive. But it is nonetheless doable, and the UN listed a number of specific ways we can achieve it. Read the full story.
How people are using GPT-4
Last week was intense for AI news, with a flood of major product releases from a number of leading companies. But one announcement outshined them all: OpenAI’s new multimodal large language model, GPT-4. William Douglas Heaven, our senior AI editor, got an exclusive preview. Read about his initial impressions.
Unlike OpenAI’s viral hit ChatGPT, which is freely accessible to the general public, GPT-4 is currently accessible only to developers. It’s still early days for the tech, and it’ll take a while for it to feed through into new products and services. Still, people are already testing its capabilities out in the open. Read about some of the most fun and interesting ways they’re doing that, from hustling up money to writing code to reducing doctors’ workloads.
Google just launched Bard, its answer to ChatGPT—and it wants you to make it better
Google has a lot riding on this launch. Microsoft partnered with OpenAI to make an aggressive play for Google’s top spot in search. Meanwhile, Google blundered straight out of the gate when it first tried to respond. In a teaser clip for Bard that the company put out in February, the chatbot was shown making a factual error. Google’s value fell by $100 billion overnight.
Google won’t share many details about how Bard works: large language models, the technology behind this wave of chatbots, have become valuable IP. But it will say that Bard is built on top of a new version of LaMDA, Google’s flagship large language model. Google says it will update Bard as the underlying tech improves. Like ChatGPT and GPT-4, Bard is fine-tuned using reinforcement learning from human feedback, a technique that trains a large language model to give more useful and less toxic responses.
Google has been working on Bard for a few months behind closed doors but says that it’s still an experiment. The company is now making the chatbot available for free to people in the US and the UK who sign up to a waitlist. These early users will help test and improve the technology. “We’ll get user feedback, and we will ramp it up over time based on that feedback,” says Google’s vice president of research, Zoubin Ghahramani. “We are mindful of all the things that can go wrong with large language models.”
But Margaret Mitchell, chief ethics scientist at AI startup Hugging Face and former co-lead of Google’s AI ethics team, is skeptical of this framing. Google has been working on LaMDA for years, she says, and she thinks pitching Bard as an experiment “is a PR trick that larger companies use to reach millions of customers while also removing themselves from accountability if anything goes wrong.”
Google wants users to think of Bard as a sidekick to Google Search, not a replacement. A button that sits below Bard’s chat widget says “Google It.” The idea is to nudge users to head to Google Search to check Bard’s answers or find out more. “It’s one of the things that help us offset limitations of the technology,” says Krawczyk.
“We really want to encourage people to actually explore other places, sort of confirm things if they’re not sure,” says Ghahramani.
This acknowledgement of Bard’s flaws has shaped the chatbot’s design in other ways, too. Users can interact with Bard only a handful of times in any given session. This is because the longer large language models engage in a single conversation, the more likely they are to go off the rails. Many of the weirder responses from Bing Chat that people have shared online emerged at the end of drawn-out exchanges, for example.
Google won’t confirm what the conversation limit will be for launch, but it will be set quite low for the initial release and adjusted depending on user feedback.
Google is also playing it safe in terms of content. Users will not be able to ask for sexually explicit, illegal, or harmful material (as judged by Google) or personal information. In my demo, Bard would not give me tips on how to make a Molotov cocktail. That’s standard for this generation of chatbot. But it would also not provide any medical information, such as how to spot signs of cancer. “Bard is not a doctor. It’s not going to give medical advice,” says Krawczyk.
Perhaps the biggest difference between Bard and ChatGPT is that Bard produces three versions of every response, which Google calls “drafts.” Users can click between them and pick the response they prefer, or mix and match between them. The aim is to remind people that Bard cannot generate perfect answers. “There’s the sense of authoritativeness when you only see one example,” says Krawczyk. “And we know there are limitations around factuality.”
How AI experts are using GPT-4
Hoffman got access to the system last summer and has since been writing up his thoughts on the different ways the AI model could be used in education, the arts, the justice system, journalism, and more. In the book, which includes copy-pasted extracts from his interactions with the system, he outlines his vision for the future of AI, uses GPT-4 as a writing assistant to get new ideas, and analyzes its answers.
A quick final word … GPT-4 is the cool new shiny toy of the moment for the AI community. There’s no denying it is a powerful assistive technology that can help us come up with ideas, condense text, explain concepts, and automate mundane tasks. That’s a welcome development, especially for white-collar knowledge workers.
However, it’s notable that OpenAI itself urges caution around use of the model and warns that it poses several safety risks, including infringing on privacy, fooling people into thinking it’s human, and generating harmful content. It also has the potential to be used for other risky behaviors we haven’t encountered yet. So by all means, get excited, but let’s not be blinded by the hype. At the moment, there is nothing stopping people from using these powerful new models to do harmful things, and nothing to hold them accountable if they do.
Chinese tech giant Baidu just released its answer to ChatGPT
So. Many. Chatbots. The latest player to enter the AI chatbot game is Chinese tech giant Baidu. Late last week, Baidu unveiled a new large language model called Ernie Bot, which can solve math questions, write marketing copy, answer questions about Chinese literature, and generate multimedia responses.
A Chinese alternative: Ernie Bot (the name stands for “Enhanced Representation from kNowledge IntEgration;” its Chinese name is 文心一言, or Wenxin Yiyan) performs particularly well on tasks specific to Chinese culture, like explaining a historical fact or writing a traditional poem. Read more from my colleague Zeyi Yang.
Even Deeper Learning
Language models may be able to “self-correct” biases—if you ask them to
Large language models are infamous for spewing toxic biases, thanks to the reams of awful human-produced content they get trained on. But if the models are large enough, they may be able to self-correct for some of these biases. Remarkably, all we might have to do is ask.