Category: 3. Business

YouTube has begun using artificial intelligence (AI) to identify children pretending to be adults on its platform, as pressure mounts on tech firms to protect minors from age-inappropriate content.

The measure is being rolled out in the United States, with YouTube and other social media platforms such as Instagram and TikTok facing scrutiny over safeguarding children online.

According to James Beser, YouTube’s director of product management for youth, a machine learning system will estimate users’ ages using factors such as viewing history and account activity.

“This technology allows us to infer a user’s age and use that signal, regardless of the account birthday, to deliver age-appropriate experiences and protections,” Beser said.

Users flagged as minors will be notified and can verify their age through a credit card, selfie, or government ID. The system builds on existing age-detection technology already in use in other markets, where it has proved effective.

The move comes as countries like Australia prepare strict social media regulations to protect children. From 10 December, children under 16 will be prohibited from accessing YouTube and other social platforms, amid concerns over harmful content and “predatory algorithms,” according to Communications Minister Anika Wells.

YouTube emphasised that it is a video-sharing platform rather than a social media network, offering a library of free, high-quality content increasingly viewed on TV screens.

The Finance Division on Friday night announced a cut of Rs12.84 per litre in the price of high-speed diesel (HSD) for the next fortnight, while petrol prices will remain unchanged.

Most of the transport sector runs on HSD and its price is considered inflationary as it is mostly used in heavy transport vehicles, trains and agricultural engines like trucks, buses, tractors, tube-wells and threshers and particularly adds to the prices of vegetables and other eatables.

A notification from the Finance Division, available with Dawn.com, stated that the new price of HSD was Rs272.99, while petrol would remain at Rs264.61. Additionally, the prices of superior kerosene oil and light diesel oil were slashed by Rs7.19 per litre and Rs8.2, respectively.

“The Government has decided to revise the petroleum product prices for the fortnight commencing August 16, 2025, in line with the recommendations of Ogra (Oil and Gas Regulatory Authority) and the concerned ministries,” the notification read.

Earlier this week, the per-litre HSD price was expected to fall by about Rs11.50, while petrol was expected to see an increase of Rs1.40 per litre for the next fortnight, driven by shifts in the global market and exchange rate movements.

International petrol prices inched up by 15 cents per barrel over the past fortnight, while diesel rates fell by about $4.5 per barrel. The rupee also appreciated slightly against the dollar.

On July 31, Prime Minister Shehbaz Sharif approved a cut in the prices of petrol by Rs6.17 per litre and high-speed diesel (HSD) by Rs10.86, respectively, for the next fortnight. The cuts in petrol and HSD prices were higher than previously estimated.

Petrol, used mainly in private cars, small vehicles, rickshaws and motorcycles, directly impacts the budgets of middle- and lower-middle-income groups.

Story Continues

Historically, the U.S. economy has displayed robust, positive productivity growth. According to data from the Bureau of Economic Analysis, labor productivity (that is, real output per worker) has more than tripled over the period 1948-2023. Moreover, this positive trend is visible in almost every broad sector of the U.S. economy.

However, there is one notable exception: construction. Labor productivity in U.S. construction in 2023 was essentially the same as it was in 1948. More importantly, as displayed in Figure 1, labor productivity in U.S. construction has actually been falling since the 1970s.

This phenomenon of falling productivity in U.S. construction was emphasized by the 2023 working paper “The Strange and Awful Path of Productivity in the U.S. Construction Sector.”¹ Real output per worker in 2020 was more than 30 percent lower than in 1970, following one of the most persistent productivity declines in any major industry.

This presents a puzzling contradiction: While other goods-producing sectors (such as mining and manufacturing) have achieved remarkable productivity gains over the past few decades, the construction sector — which is responsible for building the homes, offices and infrastructure needed for our growing economy — has experienced a prolonged and quite severe productivity decline.

This decline has significant implications not only for housing affordability but for overall economic growth. Construction represented about 4.5 percent of U.S. GDP in 2024. While this sector is less than half of manufacturing in terms of value added, the importance of a decline in construction productivity for its aggregate counterpart cannot be overstated.

The construction sector is an important supplier of capital to other sectors, so a decline in construction productivity can be amplified to the aggregate. In fact, a 2022 paper concludes that the construction sector alone accounted for about one-third of the decline in trend GDP growth since WWII.² This loss is equivalent to about $1 trillion every five years.³ Similarly, a 2023 working paper estimates that, if construction productivity had grown at even a modest 1 percent annually since 1970, annual aggregate labor productivity growth would have been roughly 0.18 percentage points higher.⁴ This difference would have resulted in current aggregate productivity — and likely income per capita — being about 10 percent higher than actual levels.

The magnitude of this decline becomes particularly striking when compared to overall economic performance. During the 1950s and 1960s, construction productivity actually grew at rates that matched or exceeded the broader economy. However, around 1970, construction productivity began a sustained downward trajectory that has continued for five decades. While aggregate U.S. labor productivity tripled between 1950 and 2020, its measure for construction remained essentially the same. This divergence is especially notable when compared with manufacturing, which experienced a large increase in labor productivity over the same period despite dealing with similar physical assembly and configuration challenges.

Is the Productivity Decline Driven by Measurement Error?

Given the overall growth of the U.S. economy and the construction sector’s apparent importance to the economy, a secular decline in U.S. construction productivity seems peculiar at first. As a result, some observers in the literature were initially skeptical about this phenomenon and believed that the drop in construction productivity was primarily driven by measurement issues.

Labor productivity requires only two components:

• Real output (that is, value added in constant dollars)
• A measure for labor

The latter is typically measured by the number of either full-time equivalent payroll workers or hours worked by these employees.

It is important to point out that construction establishments use a lot of subcontractors and/or undocumented immigrants, which are not included in traditional measures of labor. However, correcting these measurement “errors” does not reverse the declining trend in construction productivity. On the contrary, correcting productivity figures for subcontractors or undocumented immigrants would only display a larger decline in its productivity, since such a correction would mean more workers producing the same level of output. As a result, a decline in labor productivity cannot be solely caused by mismeasurement in labor variables.

Productivity Measurement and Deflators

One of the main points of controversy in the construction productivity debate lies in its deflators.⁵ When labor productivity is deflated using the GDP deflator for the whole economy, construction productivity essentially follows the economy’s upward trend, as seen in Figure 1. It is important to note that this observation does not immediately imply that commonly used deflators for the construction sector are wrong. However, increasing output prices do play a big role in understanding productivity for the construction sector.

Given that construction productivity calculations depend heavily on price deflators used to convert nominal construction spending into real output measures, the question of measurement error in these deflators has been central in the construction productivity debate. A 2025 working paper specifically investigated whether measurement problems in these deflators could account for the observed productivity decline.⁶ It focuses on potential biases arising from unobserved improvements in structure quality — such as better insulation, more energy-efficient systems and higher-quality materials — that are not fully captured in official price measurements.

Using three different analytical approaches — econometric techniques, detailed construction cost data and observable quality measures from property tax assessors and homeowner surveys — this paper finds evidence of upward bias in construction price deflators. However, the magnitude of this bias was not large enough to fundamentally alter existing conclusions about poor construction productivity performance. Its most generous estimates suggest that unobserved quality improvements may have biased construction sector productivity growth downward by at most 0.5 percentage points per year from quality-related factors, with additional small biases from other measurement issues.

Even accounting for these measurement problems, this paper concludes that “productivity growth may well have been quite low in construction, even if it has not been as low as implied by official statistics.” Construction productivity growth would still be essentially flat rather than negative, but it would remain far below other major industries: After adjusting for all potential measurement biases, construction productivity growth would still lag behind the next-lowest performing industries by about 1 percentage point annually and behind overall economic productivity growth by nearly 2 percentage points. These findings reinforce that the construction sector’s productivity challenges represent a real economic phenomenon with substantial implications for long-term U.S. economic growth.

This view is also supported by earlier studies and has been recently corroborated as well.⁷ When these studies examined physical measures of productivity (such as housing units built per worker), they found similar patterns of stagnation or decline that do not rely on potentially problematic price adjustments. Additionally, the sector has shown deteriorating efficiency in converting materials into finished products, and there appears to be limited reallocation of resources from less productive to more productive areas, which is a mechanism that typically helps drive productivity growth in well-functioning markets.

Construction Productivity Decline and Land-Use Regulation

Since the decline in construction productivity does not seem to be a mere measurement problem, a natural follow-up question is: What caused this decades-long decline in the productivity of the construction sector? An interesting hypothesis has recently been put forward by a 2024 working paper: the rise of land-use regulation.⁸

Using historical data from the Census Bureau and the Bureau of Labor Statistics, the authors construct a series on construction productivity (that is, housing units started per worker) from 1900-2023. Its analysis reveals an inverted U-shaped curve — what the authors term a “Kuznets curve” — for construction where labor productivity displayed no noticeable trend between 1900 and 1940, boomed after WWII and then sharply decreased again after 1970. This historical pattern demonstrates that technological progress in construction is not inherently impossible, making the post-1970 decline all the more striking. Then, the authors show that the timing of this productivity reversal roughly coincides with America’s embrace of increasingly restrictive land-use controls and building regulations.

The paper’s argument is relatively straightforward: Project-level building regulations systematically reduce the size of individual construction projects by making larger developments more difficult to approve through lengthy permitting processes and community opposition. Through a structural model, the paper demonstrates that, when developers face lower approval probabilities for larger projects, it is optimal for these construction firms to pursue smaller developments. In turn, this leads to smaller construction companies since these firms cannot efficiently manage numerous small projects scattered across different locations.

This regulatory constraint creates a vicious cycle that undermines productivity growth. Smaller construction firms have reduced incentives to invest in technological innovation and process improvements because the benefits of such investments cannot be spread across as many projects or housing units. Empirical evidence from the paper supports this theory: Areas with stricter land-use regulation have significantly smaller and less productive construction establishments. Furthermore, patenting activity in construction stagnated after 1970, while manufacturing patents continued to grow. In the end, the paper estimates that construction productivity could be 60 percent higher if firm sizes matched those in manufacturing. This regulatory explanation helps explain why construction has diverged so sharply from other sectors that have continued to benefit from technological progress and economies of scale.

Conclusion

The decline in construction productivity represents one of the most significant and persistent sectoral challenges facing the U.S. economy. While measurement issues may account for some of the observed decline, most evidence seems to point out that genuine productivity problems have plagued the industry for more than five decades.

The stakes of addressing this challenge extend far beyond construction companies themselves. With construction productivity problems contributing to reduced housing affordability and slower overall economic growth, finding solutions has become critical. The regulatory hypothesis offers a potentially actionable path forward, suggesting that reforms to land-use and building approval processes could help restore the technological dynamism that the sector demonstrated during its post-World War II boom.

As policymakers grapple with housing shortages and infrastructure needs, understanding and addressing the root causes of construction’s productivity stagnation will be essential for promoting U.S. productivity growth.

---

Chen Yeh is a senior economist in the Research Department at the Federal Reserve Bank of Richmond.

---

This article may be photocopied or reprinted in its entirety. Please credit the author, source, and the Federal Reserve Bank of Richmond and include the italicized statement below.

Views expressed in this article are those of the author and not necessarily those of the Federal Reserve Bank of Richmond or the Federal Reserve System.

Updated clinical data from a phase 1/2 study (NCT07020221) in China of the investigational oral KRAS G12D inhibitor, VS-7375 (GFH375), have demonstrated compelling efficacy in patients with advanced non–small cell lung cancer (NSCLC) harboring a KRAS G12D mutation.¹ The findings, which represent a significant increase in response rates from previously reported data, are scheduled for presentation at the IASLC 2025 World Conference on Lung Cancer (WCLC) in Barcelona, Spain. The data highlight the potential of this agent in a patient population with high unmet clinical need.

The study, a collaboration between Verastem Oncology and GenFleet Therapeutics, provides the latest insights into the clinical profile of VS-7375, a dual “ON/OFF” KRAS G12D inhibitor. According to the late-breaking abstract, the objective response rate (ORR) at the recommended phase 2 dose of 600 mg once daily was 68.8% (n = 11/16).

“These data at WCLC build on the encouraging data presented at [the American Society of Clinical Oncology (ASCO) Annual Meeting] earlier this year, where GenFleet initially reported an ORR of 42% in 12 patients with advanced non-small cell lung cancer harboring a KRAS G12D mutation,” said Dan Paterson, president and chief executive officer of Verastem Oncology, in a press release.

Across all dose levels in the NSCLC cohort, the ORR was 57.7% (n = 15/26), with a disease control rate (DCR) of 88.5% (n = 23/26). All patients in the NSCLC cohort had metastatic disease at baseline, with the vast majority (64.3%) having received at least 2 prior lines of systemic therapy and nearly all (96.4%) having been previously treated with an anti–PD-1/PD-L1 agent. These patient characteristics underscore the advanced and heavily pretreated nature of the study population, making the observed response rates particularly encouraging.

The phase 1/2 trial, initiated in July 2024, included a total of 142 patients across multiple tumor types,² with a median follow-up time of 4.5 months as of the July 15, 2025, data cutoff.¹ The study population was composed of patients with advanced NSCLC, pancreatic ductal adenocarcinoma (PDAC), and other solid tumors, with 28, 85, and 29 patients, respectively. While efficacy data were specified for the NSCLC cohort, the safety profile was reported cumulatively across all 142 patients. The safety analysis revealed that treatment-related adverse events (TRAEs) were predominantly grade 1 or 2 in severity. The most common TRAEs, occurring in at least 20% of patients, included diarrhea, vomiting, nausea, and decreased appetite. The incidence of grade 3 or higher TRAEs and severe AEs was 27.5% and 7.7%, respectively. Of the 142 patients in the safety population, 11 required dose reduction and 6 discontinued treatment due to TRAEs, with no TRAE-related deaths reported. These data suggest a manageable safety profile for VS-7375 at the doses studied.

The KRAS G12D mutation is the most prevalent KRAS mutation, accounting for 26% of all KRAS-mutated cancers. It is particularly common in pancreatic cancer (37%) and colorectal cancer (12.5%), as well as NSCLC (5%). Despite its prevalence, there are currently no FDA-approved therapies specifically targeting this mutation. The dual “ON/OFF” mechanism of VS-7375 represents a novel approach to inhibiting both the active and inactive states of the KRAS G12D protein, distinguishing it from other agents that may only target 1 state.

In April 2025, the FDA cleared the investigational new drug application of VS-7375.³ Last month, the FDA granted fast track designation to VS-7375 in KRAS G12D-mutated PDAC.⁴

REFERENCES:

1. Verastem Oncology Announces Late-Breaking Abstract from Partner GenFleet Therapeutics’ Study in China of GFH375 (VS-7375) in Advanced Non-Small Cell Lung Cancer at IASLC 2025 World Conference on Lung Cancer. News release. August 13, 2025. Accessed August 15, 2025. https://tinyurl.com/mr24mmuk

2. A Phase 1/​2a Study of VS-7375 in Patients With KRAS G12D-Mutated Solid Tumors. ClinicalTrials.gov. Updated July 18, 2025. Accessed August 15, 2025. https://clinicaltrials.gov/study/NCT07020221

3. Verastem Oncology announces US IND clearance of VS-7375, oral KRAS G12D (ON/OFF) inhibitor, enabling phase 1/2a trial in advanced solid tumors. News release. Verastem Oncology. April 24, 2025. Accessed August 15, 2025. https://tinyurl.com/3db7zt28

4. Verastem Oncology Granted Fast Track Designation for VS-7375 for the Treatment of KRAS G12D-mutated Locally Advanced or Metastatic Pancreatic Cancer. News release. Verastem Oncology. July 24, 2025. Accessed August 15, 2025. https://tinyurl.com/5e7ftfd9

By Genna Contino

The pandemic changed dating for Generation Z. Economic pressures are making it even harder.

As the cost of living continues to rise and pandemic-fueled shifts in how people socialize linger, more than half of Gen Z is spending $0 monthly on dating, according to a Bank of America report.

For a demographic cohort facing a high cost of living, mounting student-loan debt and job-market instability, Generation Z has not been placing a high priority on dating. Instead, it’s a financial burden.

Data on spending habits provide a glimpse of how young folks’ mindset toward dating has shifted. About half of men (53%) and women (54%) ages 18 to 28 are spending $0 a month on dating, a recent Bank of America (BAC) report on Gen Z adults’ financial health found.

This is likely because Gen Z has a different approach to dating compared with prior generations when their members were in the age bracket Gen Z is in now, experts say. With pandemic-fueled changes to how people socialize, alongside economic pressures fueling a pullback in young people’s spending, dating has taken a backseat for a significant portion of Gen Z. Instead, many are prioritizing economic security before rushing to find a partner, get married, have kids and buy a home, compared with past generations, who typically hit those milestones earlier in life.

See more: TikTok’s latest budget hack is ‘dating for dinner’ – which says a lot about love and money right now

The Gen Z dating scene has been shaped by dating apps, social media and generative AI in ways that previous generations did not experience, while a period of forced isolation during the coronavirus pandemic exacerbated young people’s reliance on online connection, which in turn altered the way they connect. A 2024 Hinge report found that Gen Z is 50% more likely than millennials to wait to answer a message from a potential match to avoid seeming overzealous, and more than half (56%) surveyed said the fear of rejection has kept them from pursuing a potential relationship.

“We’ve seen a continual decline in dating. Some people are calling this the ‘romance recession’ or the ‘sex recession,’ particularly when we look at Gen Zs and you think of when they came of age, and dating and the pandemic hit within that,” said dating coach Damona Hoffman. “I see more people tapping out of the dating market entirely and just having anxiety about even stepping into the dating scene if they don’t have the money to date.”

Why is half of Gen Z not spending on dating?

About half of the Gen Z members surveyed by Bank of America said the high cost of living is a barrier to financial success. “They’re worried about groceries, worried about rent, worried about dining out,” said Will Smayda, head of financial centers at Bank of America.

And prices are still on the rise. At the core of the consumer-price index, an inflation reading that excludes food and energy prices and is a better predictor, prices were up 0.3% in July – the sharpest monthly increase since January.

See: Key inflation rate shows biggest rise in 6 months, CPI shows, but Fed rate cut still appears in play

Also: Americans more worried about inflation and unemployment after new trade deals leave tariffs at highest level in decades

Experts agree these growing financial hardships are making it hard for Gen Z to prioritize dating. And when members of the generation do decide to spend time and money on a date, they’re more deliberate with their choices.

“We’re seeing a shift toward intentional dating, especially among younger singles,” said Michael Kaye, director of brand marketing and communications for the dating app OkCupid. “Gen Z isn’t matching for instant gratification; they’re dating with purpose and prioritizing compatibility and shared values, over flashy spending.”

Some Gen Z-ers have chosen to tap out of the dating scene because of their own insecurities. Hoffman said she often comes across Gen Z singles who feel they are not “dateable” because of debt or a low-paying job, and a fear of being judged by potential partners keeps them from dating altogether.

However, data suggest that these fears may be unfounded: 68% of singles surveyed by OkCupid said they would date someone who makes significantly less than them, provided that person is driven or money is not a priority for them in a relationship. Gen Z and millennials made up 90% of the survey respondents.

People anxious about their own financial status and career might not realize that much of the dating pool is in the same boat, Hoffman said. And even if people think or say they’re looking for a partner who is rich, the person they end up connecting with might not fit that bill.

“What they say they want on a dating app or what they are kind of aspirationally going for in a match is not necessarily all they will date,” Hoffman said.

See more: Welcome to the real world, Class of 2025. Here are 5 money tips you haven’t heard a million times.

These money concerns tend to loom larger for singles when a relationship is still hypothetical, Hoffman added. “When you actually make that connection and they find some commonality, the financial piece usually is not front-loaded as much as it is when somebody is thinking about dating or planning to date.”

While these various stressors haven’t kept the entire generation from dating, they have made those who are dating more budget-conscious. Just under half of those surveyed by Bank of America said they are spending money on dating, with 28% of the total survey sample saying they spend less than $100 a month. Of the Gen Z-ers in a relationship, 43% of those not living together spent anywhere from $1 to $100 a month on dates, and 48% of cohabitating couples also fell in that range.

Gen Z’s postpandemic dating is different from generations prior

Along with financial pressures, the COVID-19 pandemic accelerated a trend of young people turning to online spaces for connection. While this sometimes meant engaging in solitary activities like binge-watching TV shows, it also led to a greater reliance on online communities to maintain social ties during quarantine. A Pew Research Center study, for example, found that adults under 30 were likeliest to say social-media sites helped them stay connected to friends and family during the pandemic.

For many young people who wanted to be out and about socializing and dating, this online connection was a lifeline. However, it created a stark contrast with previous generations, which relied on in-person gatherings and phone calls to form connections, using social media as a supplement to socializing rather than as a replacement, the Pew study suggests.

From the archives (April 2020): Online dating amid coronavirus: Longer conversations and a ‘pivot’ to video dates

Gen Z is also the first generation to come of age during the era of generative artificial intelligence, and while generative-AI companions are growing in popularity, their impact on dating is nuanced. A third of Gen Z singles have engaged with AI as a romantic companion, compared with 16% of all American singles, a 2025 study by Match Group (MTCH) in collaboration with the Kinsey Institute found.

“If you have Chad on your computer for free, why are you going to pay money to meet a real person who might reject you?” Hoffman said.

This trend toward AI companionship among singles, while notable, does not provide a full picture of Gen Z’s relationship with the technology. A broader 2024 Pew Research Center survey found that the majority of adults under 30 are mostly using generative AI tools for work, learning and entertainment, rather than for romantic purposes. So while the role of AI in dating is growing, research suggests it is not a mainstream replacement for human relationships.

Nonetheless, as Gen Z navigates evolving societal and economic pressures, a ripple effect is appearing: Dating is being delayed, which then pushes back major milestones such as marriage and homeownership.

Young adults are prioritizing economic security over having kids, Census Bureau data show, reflecting the rising cost of living. The S&P CoreLogic Case-Shiller national home-price index, which tracks U.S. residential real-estate prices, has nearly doubled over the past decade, while annual prices for child care ranged between $6,500 and $15,600 in 2022. The cost of tuition at a four-year public university increased 36.7% between 2010 and 2023, according to the Education Data Initiative.

See more: Home prices post smallest increase in nearly two years, Case-Shiller index says, offering buyers a reprieve

These rising prices have served as barriers for young people looking to settle down. In 1975, 45% of people ages 25 to 34 had moved out of the house, gotten married, had children and bought a home. In 2024, just 21% of people in this age group had reached all four of these milestones.

“Their milestones are not the same as their parents’ milestones,” Hoffman said. “The goalposts have moved.”

See: Biggest wholesale-prices jump in three years is clearest sign yet that Trump’s tariffs are costing Americans

Reframing a budget can help take the pressure off dating

One promising finding from the Bank of America report about Gen Z’s financial health is that 72% of the young people surveyed took steps to improve their financial health over the past year, including putting money toward savings (51%) or paying down debt (24%).

Gen Z-ers are also more straightforward than their older counterparts about their financial situation. They’re more likely to participate in “loud budgeting,” or openly share their budget strategies, said Smayda of Bank of America, and 42% in the report said they feel comfortable turning down a social outing if they can’t afford it.

From the archives (January 2024): ‘Loud budgeting’ is in for 2024, TikTok users say. It’s kind of a joke – but experts say it could help you.

(MORE TO FOLLOW) Dow Jones Newswires

08-15-25 1411ET

Copyright (c) 2025 Dow Jones & Company, Inc.