Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Hold tight to your free weights — the Ozempic revolution is coming to a gym near you.

The runaway success of “GLP-1” weight loss and diabetes drugs, which also include Wegovy and Zepbound, is hard to overstate. Sales are expected to approach $50bn this year, making them the top-selling class of drugs worldwide. That is despite global shortages, high prices and the fact that the drugs are largely available only in injectable form so far. Sales are expected to more than double to $130bn by 2030 and could soar higher if the makers win permission to sell them as a preventive tool.

For pharma groups Novo Nordisk and Eli Lilly, soon to be joined by others, this is fabulous news. For others, it is likely to be really bad. Diet company WeightWatchers recently changed chief executives as it struggles to adjust, and soft drink, beer and snack company shares have been on a wild ride as investors try to figure out who will be hurt the most when consumers taking the drugs eat healthier food and fewer calories overall.

For gyms and health clubs, the impact is going to be huge but complicated for an industry that is still rebuilding after Covid. The pandemic put a quarter of US fitness centres out of business and reshaped commuting and exercise patterns. Weight-loss drugs are likely to supercharge a consumer rush towards strength training equipment that has been gaining force for more than a decade, and many gyms are still ill-prepared.

Ten years ago, most health club floors were seas of treadmills, elliptical machines and stationary bikes, with fixed weight machines along the edges along with a free weight area geared towards power lifting, mostly by men. But the pandemic and concurrent rise of apps and YouTube videos that gave people access to personalised fitness routines has made that configuration all but obsolete.

Customers still use treadmills but both sexes now seek out a wider range of strength training equipment, including barbells, dumbbells, medicine balls and the like. Clubs, seeking to boost membership, have also leaned into the social aspects of in-person fitness, from group classes and personal trainers to cafés and hang-out areas.

Gyms are pushing their stair climbers and fixed weight machines to the periphery and replacing them with open space for body-sculpting classes, free weights and individual training sessions.

“We’re seeing a greater demand for space for strength,” Colleen Keating, CEO of Planet Fitness, one of the largest listed gym groups, told analysts in August. Even Peloton, famous for its cardio-intensive bikes, is testing an app focused on strength training.

The shift takes time and money. The now less-popular cardio machines are often sold on multiyear leases, while strength training equipment generally requires an upfront investment. The delay is leading to uneven usage and customer complaints at clubs that have not made the shift.

Weight-loss drugs will exacerbate the pressure. As the drugs gain acceptance, fewer people are likely to rely on exercise as their primary weight loss tool and the drugs’ side effects, nausea and intestinal distress, can make high-impact cardio activities uncomfortable.

However, GLP-1 users still need the gym. Studies suggest that the drugs cause significant muscle loss along with fat, leading to problems with balance and mobility as well as saggy skin sometimes dubbed “Ozempic butt”.

Strength training seems to be the answer not just for GLP-1 users but everyone else. A growing body of medical literature suggests strength training cuts mortality, particularly for women, while also helping to prevent osteoporosis and relieving the symptoms of depression.

“It’s gone from being health and fitness to health and wellness, which is a lot more holistic” says Eleanor Scott, a partner on PwC’s leisure strategy team.

Foot traffic to popular US gym chains Crunch Fitness and EoS Fitness is up by double-digits year on year, according to data provider Plaicer.ai. Planet Fitness has added 2.7mn members since the start of 2023 and improved its profit margins.

For all of them, the combination of strength training with prevention creates a chance to win, or win back, older customers still wary of gyms post-Covid. Although 80 per cent of baby boomers participate in fitness activities, just 42 per cent belong to a gym, compared to nearly three-quarters of active Gen Zers and millennials, according to ABC Fitness. But growth will not follow if newcomers end up fighting the regulars for access to the dumbbells.

brooke.masters@ft.com

Follow Brooke Masters with myFT and on X

(Reuters) – Russia will run a nationwide test of its emergency public warning systems on Wednesday, letting sirens wail and interrupting television and radio broadcasts in a twice-yearly initiative amid the war in Ukraine.

At around 10.30 a.m. in most of Russia’s 11 time zones, sirens will sound for a minute, with loudspeakers broadcasting an “Attention everyone!” call, the emergency ministry said on the Telegram messaging app.

The exercise aims to check the warning systems, the readiness of those responsible for launching them and to raise public awareness, the ministry said, adding, “Don’t panic – everything is according to a plan.”

The frequency of rehearsals was doubled from last year, following the first event held in 2020.

It comes amid Russia’s war in Ukraine, which Moscow started in 2022, triggering the deepest crisis in its relations with the West since the Cuban Missile Crisis of 1962.

(Reporting by Lidia Kelly in Melbourne; Editing by Clarence Fernandez)

Ajesh Kumar, a college graduate in a village in Haryana, a rural state bordering Delhi, recently applied to work as a cleaner. But there were more than 400,000 jobseekers for an estimated 5,000 positions, making the 30-year old’s chances about one in 80. 

“There’s just no hope, no chance” of getting one of the government posts, Kumar said, which are prized because of the guaranteed hours, wages and benefits, however low, of public sector work. Among the applicants were two of his family members.

Kumar is one face of India’s most intractable public policy issue: a chronic shortage of formal jobs in the world’s most populous country and, according to companies, a corresponding shortage of suitable candidates to fill them.

Prime Minister Narendra Modi’s economic record will again be on the agenda in Haryana on Saturday in one in a string of regional polls in which the opposition will seek to build momentum against his Bharatiya Janata party. The opposition managed to push the BJP into a parliamentary minority for the first time since 2014 in nationwide elections this year, in part by highlighting persistently high joblessness.

India’s economy is failing to create enough jobs for its young and growing population and train the skilled workers its companies need to harness that demographic dividend. This mismatch is feeding widespread grievances and represents one of the biggest challenges for Modi as he enters his second decade in power.

“Every month about a million formal job seekers are being added to the workforce,” says Rituparna Chakraborty, co-founder of Teamlease, which describes itself as India’s biggest staffing company. “Nine out of 10 of them go into the informal sector — jobs where there is no employment contract, no social security benefits, no protection, and no wage guarantees.” 

“The poorest Indians tend to take on daily wage jobs in things like construction because there aren’t too many alternatives,” says Shruti Rajagopalan, senior research fellow at the Mercatus Center at George Mason University, in Virginia.

“The people in the middle are still waiting, and would rather hold out for a government job, or work on the family farm because at least it provides them food security.” 

Modi’s government has taken steps to tackle India’s joblessness. In the first post-election budget, finance minister Nirmala Sitharaman announced an apprenticeship scheme aimed at benefiting 10mn young people over five years. The government has also promised training subsidies for companies, stipends for apprenticeships and help for vocational schools to amend their curricula to align with job market demands. 

In its previous term, Modi’s cabinet also cut corporate taxes and took steps to amend labour laws in a bid to stimulate job growth.

Corporate India, however, laments a shortage of qualified candidates for its top jobs. Conglomerate Larsen & Toubro said in June that it faced a shortage of 45,000 skilled labourers and engineers across its businesses, which range from construction to information technology.

Analysts said the skills gap bodes ill for Modi’s “Make in India” manufacturing push, and attests to neglect and uneven standards at Indian secondary institutions.

“So many people come out of these colleges, but we can do a lot to make them more employable in the industry,” K Krithivasan, chief executive of Tata Consultancy Services, India’s biggest IT company, told the Financial Times earlier this year. 

Mohandas Pai, chair of private equity firm Aarin Capital and former chief financial officer at IT giant Infosys, said most industries were struggling to find skilled workers as India’s economy expands at an annual clip of about 7 per cent, with job openings outpacing the supply of employable workers.

At the same time, he said: “Many industries are not willing to spend money to hire them, skill them and train them.” 

A study published this year by Quess Corp, an Indian business service provider, and the Federation of Indian Chambers of Commerce and Industry argued that India faced a wage — rather than an employment — problem. About 80 per cent of jobs pay less than Rs20,000 ($238) a month, not enough to meet rising living expenses, the study’s authors argued. 

On the supply side, economists say cumbersome labour regulation is also holding back industry from creating jobs. Much of the legislation only kicks in for companies employing 10 people or more, points out George Mason University’s Rajagopalan. “Either people are not hiring the 10th worker, or they hire the worker informally,” she said.

Modi’s government in 2020 approved an overhaul of India’s patchwork of labour laws, which regulate areas ranging from maximum shift hours to the number of clocks per factory floor. But the reforms have yet to take effect. 

There is even disagreement over how to measure India’s unemployment. The Centre for Monitoring Indian Economy, a think-tank, publishes the most widely cited indicator, which is conducted monthly. In August, it showed a jobless rate at 8.51 per cent, and unemployment on a rising trend. 

“This is a pretty high unemployment rate in a country growing at 7 to 8 per cent per annum,” says Mahesh Vyas, CMIE’s managing director. “We have also been seeing the unemployment rate very high for a long period now in both rural and urban regions.” 

Modi’s political circle favours the Periodic Labour Force Survey, which reports quarterly rural and urban unemployment rates and shows the jobless rate at below 5 per cent and falling.

Analysts said the discrepancy was because of what counted as work, including part-time agricultural work.

Vyas claims the definition of a job in the PLFS is “too relaxed”. He also pointed to growth of India’s net fixed assets at companies, which he said served as a proxy for employment and joblessness and has been growing at only about 5-6 per cent in recent years.

“Employment will increase only if investments increase, and I don’t see that,” Vyas said.

Kumar, in Haryana, for example, might or might not qualify as unemployed depending on who is counting. He is earning some money on commission for a company that sells cattle feed, and is considering setting up a dairy business with his brother.

Like many young Indians, he also aspired to an army post, completing a correspondence degree in political science and passing the written test three times. But he was rejected in the interview.

“You need sources and contacts when you reach that level,” Kumar said. “I did not have them.”

“I have given up looking for jobs,” he added.

There’s no escaping artificial intelligence (AI), with many new solutions on the market for advisers.

But among the myriad messages, it’s rarely clear what type of AI is being promoted and whether it is suited to your needs.

There are two main types of AI today and it’s important to distinguish which one is being offered to you.

Before you commit, you must really consider what you want the tech to do for you.

Predictive AI

Often considered ‘traditional’ AI, this class of machine learning is trained to recognise patterns in data, text or speech.

Humans (data-annotation specialists) manually mark up data records with what they mean, and data scientists feed this training data into a ‘model’ – a statistical engine they have designed, usually based on some form of neural network technology.

Generative AI is only useful when a human is iteratively interacting and checking every response

This model is then used on new, previously unseen, data to predict what it should be labelled as.

Predictive AI can be used to distinguish whether a picture is a cat or a dog, or if a sentence concerns a client’s financial objectives, their emergency funds or their attitude to risk.

It can be made very accurate and, more importantly, it can be tested as to its accuracy. Its results are repeatable and any biases in its training can be removed.

In short, predictive AI is very good if the time is taken to train it well.

Generative AI

This is the ‘new’ AI – like ChatGPT, or at least the underlying GPT models from OpenAI, Facebook, Mistral, Baidu, Anthropic, Google and others.

You’ll often see these models described as large language models – or LLMs – although that’s a misused coinage, as there are many LLMs in the predictive AI family.

When buying any AI product, be explicit about what you want it for and ask each vendor to explain why their method is best suited to that

The clue to the purpose of generative AI is in the name. This class of AI is designed to generate new data, such as pictures, prose or speech.

It has no capability to understand or analyse your data, merely creating new content based on instructions or prompts. For example, if given an instruction like ‘draw me a parrot’ or ‘write a poem about the sea’ it will do so.

In simple terms, it does this by generating a likely start word from a limited random selection, picking a good next word that is statistically likely, then a third word and so on.

It doesn’t know what it is saying. It simply churns out a sequence of words statistically related to the prompt provided, based on what it has seen before – the data the GPT vendors have trained it on, mostly large portions of the internet.

So, generative AI is good at creating content, whereas predictive AI is good at identifying content.

The confusion

Many vendors are promoting generative AI that appears to understand or identify content.

In reality, they are first performing a simple search into your content to try and find relevant information, then using the first few search results within their GPT query prompt in order to formulate an answer.

It’s rarely clear what type of AI is being promoted and whether it is suited to your needs

Vendors use generative AI as a shortcut to painstakingly labelling data and training a predictive AI model suited to your needs.

The problem with this is that the search request itself is automatically generated, then only a handful of findings are used in generating an answer. The answer is then based on the standard GPT method of statistically generating one word at a time.

That’s plenty of opportunities for errors to creep in, inconsistencies to arise and even hallucinations to appear. In short, you can’t rely on the outcomes (not verified by a human) if you want to use this information for any kind of decision making.

Generative AI is only useful when a human is iteratively interacting and checking every response, such as in a chat or search application.

If you need reliable AI that will consistently identify relevant content or what it means, there is no shortcut to using predictive AI, especially if you want to limit the need for humans to check every answer.

When buying any AI product, be explicit about what you want to use it for and ensure you ask each vendor to explain why their chosen method is best suited to that and, most importantly, how they can guarantee its accuracy and data reliability.

Joe Norburn is chief executive at TCC Group and Recordsure

1.9K

By: Steve Macek

In January 2024, top executives at X (formerly Twitter), Meta (parent company of Facebook and Instagram), Snap, Discord, and TikTok appeared at a Senate hearing to answer questions about protecting children and teens online. In attendance were parents whose children had been harmed by or died as a result of their social media consumption. A climactic moment in the hearing came when Meta CEO Mark Zuckerberg apologized to the parents in the audience.

There is reason to suspect that social media use may be connected in some way to increasing depression and anxiety in young people and adults alike. Indeed, in May 2023, the US Surgeon General issued a health advisory warning that teens who use social media for more than three hours a day put their mental health at risk. But the nature of the connection between social media use and mental illness is murky at best and there is plenty of research that suggests no connection whatsoever.

Moreover, concerned parents and members of Congress have seized on concerns about social media to push an ill-considered piece of internet censorship legislation called the Kids Online Safety Act (KOSA) that will not make children any safer online, but will violate the free speech rights of both young people and adults, and will likely be weaponized against LGBTQ+ youth. As Project Censored has previously reported, legislators have “capitalized on the moral panic surrounding the impact of social media to propose problematic legislation” such as KOSA, resulting in a wave of legislation aimed at protecting children online.

The Kids Online Safety Act
Sponsored by Senators Richard Blumenthal (D-CT) and Marsha Blackburn (R-TN), KOSA was first introduced in 2022 following Facebook whistleblower Frances Haugen’s explosive revelations about the damaging impact of Instagram use on mental health in teens and children. The bill would impose a “duty of care” on websites and apps—principally, social media apps—to “prevent and mitigate” harms to children, such as anxiety, depression, suicidal ideation, drug abuse, and eating disorders, associated with their services.

It would require platforms to take steps to avoid recommending content that might promote mental health disorders to minors and ban advertising of age-restricted services (such as online gambling) and products (alcohol, tobacco) targeted at them. Under the legislation, platforms would have to give minors safeguards they can use to limit communication with others, restrict access to their private information, and protect their geolocation data. KOSA would also mandate that social media platforms provide parents with tools to protect their children’s safety online and require that they be notified if their children are exposed to potentially harmful materials.

In the version of the bill being considered last year, the legislation charged individual state attorneys general with enforcing its “duty of care” provisions, potentially allowing right-wing attorneys general in states like Florida and Texas, who are busy banning books with LGBTQ+ characters and prosecuting abortion providers, to decide what is “harmful to minors.” In response to criticisms from civil libertarians, the bill was revised so that, in its current iteration, the Federal Trade Commission has responsibility for enforcing the “duty of care” provision of the act while individual state attorneys general will be expected to enforce its safeguards for minors, transparency, and reporting requirements.

The legislation enjoys broad, bipartisan support. More than 60 senators, including Senate Majority Leader Chuck Schumer (D-NY), have endorsed the bill. A coalition of nonprofits, children’s advocates, and civic organizations, including the American Psychological Association, the American Academy of Pediatrics, the National Education Association, and the Eating Disorders Coalition, have urged its passage. And President Joe Biden has said he supports the law.

Moreover, polls indicate that huge majorities of American voters approve of government action to mitigate the harms caused by social media and favor most of the key components of the Kids Online Safety Act.

KOSA is a Censorship Bill That Will Undermine Young People’s Privacy
Although certain provisions of the Kids Online Safety Act—such as its prohibition on features of social media platforms designed to encourage compulsive user behavior—might be defensible, at its core the bill represents a massive expansion of government censorship that will cut young people off from legal and, in some cases, potentially life-saving content.

The 2023 version of KOSA rightly alarmed many LGBTQ+ rights advocates, who worried that it would be weaponized against gay, lesbian, queer, and trans youth. They were especially concerned that conservative state attorneys general might exploit the authority given them by the legislation to pressure websites into removing information about gender-affirming care or block young people’s access to LGBTQ+ online communities. Indeed, one of the bill’s main co-sponsors Marsha Blackburn said in a speech that one of the bill’s top priorities is to shield children from “the transgender in this culture.” The conservative Heritage Foundation also celebrated KOSA as a means to guard children “against the harms of sexual and transgender content.”

Though in its current version KOSA will be enforced by the FTC and not state attorneys general, the bill’s “duty of care” will still, likely, lead platforms and websites to prohibit any discussion of subjects that might get them into trouble. Educational content about sex, abortion, contraception, LGBTQ+ identity, depression, eating disorders, suicide, and other sensitive topics will simply disappear from the web or be placed behind age-verification walls. As Evan Greer of Fight for Freedom explains, “The way that this bill would work, it would just suppress all discussion of eating disorders among young people, because at scale, a platform like YouTube or Instagram is not going to be able to make a meaningful determination between, for example, a video that’s harmful in promoting eating disorders, or a video where a young person is just speaking about their experience with an eating disorder.” Being cut off from supportive online communities and information about their gender identities, sexuality, and health will hurt, rather than help, young people, and will be particularly devastating for young people from minoritized and marginalized groups.

Though one concern fueling support for KOSA is that social media companies rampantly violate the privacy of children and teens who use their platforms, the bill will actually encourage platforms and websites to further compromise users’ privacy by pushing them to adopt some sort of “age verification” scheme. While it is true that KOSA does not explicitly require age verification, websites and apps will have no choice but to require users to submit a government-issued identification or undergo biometric screening as a condition for accessing their services. How else will they be able to distinguish adult users from minors and avoid liability? Naturally, age verification will scare many adult users off the platforms that implement it. Moreover, an age verification scheme would eliminate whatever anonymity that users have online, undermining their First Amendment right to anonymous speech.

Lastly, it is worth noting that well-informed critics with expertise on the subject point out that KOSA is one of several bills that threaten the use of end-to-end encryption, which provides privacy protections that benefit all kinds of users, but especially members of marginalized communities.

Alternatives to KOSA
Despite the fact that KOSA appears poised for victory in the Senate, all is not lost. The proposed legislation is opposed by several civil liberties advocates and tech freedom groups, including the Electronic Frontier Foundation, the American Civil Liberties Union, the Woodhull Freedom Foundation, and Fight for the Future.

These groups and other critics of the bill have suggested a number of ways the government could protect the health and safety of young people online without resorting to the sort of censorship and surveillance incentivized by KOSA.

To begin with, the government could implement increased data privacy protections for all internet users. As EFF, among others, has argued, Congress could explicitly prohibit the widespread practice of “surveillance advertising,” which exploits data about users’ online behavior to target them with specially-tailored ads. Indeed, the Biden administration need not wait for Congress to act as the FTC has already created rules to protect children from “surveillance ads” and has the power to eliminate this form of ad tech altogether if it so desires.

In addition, the US could adopt some version of Europe’s General Data Protection Regulation (GDRU), requiring transparency about the data that websites gather about users and giving them a legal right to request any non-newsworthy personal information held by sites or apps be erased. Already five states—California, Colorado, Connecticut, Utah, and Virginia—have adopted GDRU-inspired data privacy laws. A federal law creating a right to control one’s personal data and a “right to erasure” of that data ought to be considered.

Greater transparency about, and systematic auditing of, the algorithms and AI that determine the online experience for both adults and minors would also help address some of the concerns Congress has about online media’s influence over children.

Finally, rather than hoping without evidence that censoring Instagram or TikTok will somehow ameliorate the country’s epidemic levels of depression and anxiety, we ought to instead fully fund public mental health services for all Americans. Even mainstream politicians acknowledge that more funding and resources are needed to address the country’s current mental health crisis. The poorest areas of the country tend to have the highest levels of depression. So, funneling more money for treatment into these high-needs areas should be a top priority.

Despite its bipartisan support, KOSA does a disservice to the young people it ostensibly aims to protect, by violating their First Amendment right to anonymous speech, potentially cutting off their access to perfectly legal content, and undermining important privacy protections that benefit all internet users. The government should impose more regulations on the way big tech companies and social media platforms harvest and use data from children and adults alike. But the sort of censorship proposed by the Kids Online Safety Act is not the answer.