July 2023 Issue of the Employment in New York State Newsletter

View as a webpage  /  Share

by Elena Volovelsky, Labor Market Analyst – New York City Region and Kevin Alexander, Labor Market Analyst – Capital Region

Most of us have daily interactions with programs that perform routine tasks like checking our spelling, sending out automated emails or responding to voice prompts over the telephone. This set of computer-based, decision-making programs — all of which follow an explicit set of business rules and procedures — is collectively known as “traditional” Artificial Intelligence (AI).

The entire AI field was upended in November 2022, when the San Francisco-based firm OpenAI announced it had developed a new program that could imitate human conversations and generate images and video called ChatGPT. Unlike traditional AI programs, ChatGPT is an example of “generative” AI (GenAI), which uses “deep machine learning” to determine the rules after being shown many examples, such as the enormous quantity of text available on the Internet, to generate entirely new content. In March 2023, Google introduced its own conversational GenAI chatbot, named “Bard.”

Generative AI’s Economic Potential

GenAI expert Erik Brynjolfsson of MIT recently examined the potential impact of this technology on productivity—the primary driver of long-term growth and prosperity—in the U.S. over the next 20 years. Productivity in the U.S., which more than doubled between the end of WWII and the early 1970's due to the automation of manufacturing processes, has slowed dramatically in recent decades, hampering wage growth and economic growth.

However, new GenAI programs have the potential to reboot economic growth with their ability to assist with tasks that used to be reserved for human intellectual work. Brynjolfsson estimated that if workers become more efficient due to assistance from AI, that alone has the potential to raise productivity by half compared with the current government forecast. And if AI assistance induces further innovation, productivity could nearly double.

A recent report from McKinsey found that most of GenAI’s economic contribution will come from boosting productivity by helping workers to automate tasks in several areas, including customer operations, sales, software engineering, and R&D. Similarly, a recent New York Times article outlined the four main types of GenAI products firms are now purchasing:

• Generate code for software engineers
• Create new content such as sales emails and product descriptions for marketing teams
• Search company data to answer employee questions
• Summarize meeting notes and lengthy documents

Rapid advancements in GenAI have attracted many start-up companies to the field. Data from CB Insights, a technology research firm, indicate there have been 91 deals in 2023 so far, which have already attracted $14.1 billion in equity funding, including a $10 billion investment by Microsoft in OpenAI. Even excluding the OpenAI deal, that represents a 38% increase from full-year 2022 levels.

Real-world AI

In a real-world example of human-machine collaboration, a large call center operator found that deploying an AI language bot to assist customer service agents made them significantly more effective in resolving problems. Assisted by AI, workers were able to resolve customer problems 30% faster and handle more queries per hour. In addition, customer satisfaction was higher, which resulted in lower turnover among employees, saving the company money.

One view of the coming changes in the job market is that workers who currently lack expertise in a subject—writing and math, for example—will become more productive with the help of these new AI tools and will be able to qualify for better-paid jobs. In other words, jobs that are currently reserved for higher-skilled workers could be done by lower-skilled workers with assistance from AI. That, in turn, could open new employment opportunities for more people without a college degree.

Current State of AI Adoption

In 2019, the U.S. Census Bureau introduced a new module into its Annual Business Survey that sought to measure adoption of advanced technologies by firms. Initial results —based on the 2016-18 period—showed that AI usage was concentrated in larger, newer companies. While only one firm in 30 consistently used AI, nearly 13% of workers were exposed to it because their employers represented larger firms. Unsurprisingly, firms in the information sector used AI the most, followed by professional services and finance.

More recently, consulting company Accenture released a report that attempted to quantify how the emerging capabilities of AI language tools might impact individual industries and occupations. They identified 200 occupational tasks related to language and measured the share of these tasks performed within major industries.

According to the Accenture report, about 30% of all working hours could be highly impacted by language AI tools, with another 10% affected by augmentation. The finance, information and professional services sectors led the pack on potential exposure, which tallied with the data from the Census survey.

OpenAI also attempted to estimate the potential workforce impact of generative AI. Through a combination and comparison of human-produced and entirely machine-generated statistical models, OpenAI estimated that four-fifths of the U.S. workforce could have 10% or more of their jobs performed by AI. A quarter of these workers could see 50% or more of their work tasks either augmented or taken over by GenAI. Jobs that require repetitive tasks, a superficial level of data analysis and routine decision making were found to have the highest risk of exposure.

Yet another measure of AI potential comes from the IBM Global AI adoption index, which surveys businesses worldwide. Last year, their survey showed that a quarter of U.S. respondents use AI in some capacity and another 43% are exploring its use. The top factors driving AI adoption were cost savings, ease of use and ready availability of AI applications. Significantly, about a fifth of all business also cited skill and labor shortages as a contributing factor.

Ethical Considerations

While the output GenAI produces can be extremely convincing, the information is not guaranteed to be correct or unbiased. GenAI models are often trained on data collected from the Internet, which may have incorrect or biased information. The output of these models can also be manipulated to enable unethical or criminal activity. Organizations that use GenAI models need to consider reputational and legal risks involved in publishing and utilizing—however unintentionally—the material that is biased, offensive or copyrighted.

In fact, the New York City Council passed a law which took effect in July 2023, making it unlawful for NYC employers to use an AI-based decision tool to screen a candidate or an employee unless that tool undergoes a bias audit. The results must be made publicly available on the company’s website.

Various states have also begun to consider ways to regulate the use of AI, including during the hiring process. Just this year, nine states have considered laws that would impose safeguards for the use of AI tools. Although it’s unlikely that any of these bills will pass this year, the increasing interest in AI will certainly be the subject of future legislation.

Final Thoughts

The debate over how GenAI tools like ChatGPT will affect jobs and the overall economy is just beginning to take shape. While some experts predict that AI will displace workers, others believe these new tools can greatly increase worker productivity and thereby boost the economy. Clearly, the AI landscape is fast-moving, but it will take some time to see how these predictions pan out.

by Elena Volovelsky, Labor Market Analyst, New York City Region

“The office market matters to the rest of New York City.”

By any measure, New York City is a global business powerhouse. More than 250,000 enterprises are located here, including the headquarters of 43 companies in the S&P 500 Index, the most of any U.S. city. These and other companies have a significant impact on economic activity and employ large numbers of professionals, support personnel and maintenance staff.

Before the COVID-19 pandemic, these companies also needed large shares of office space, thus creating the world’s largest office property market. In fact, 2019 was the second-best year on record for New York City’s office market. Sales jumped 30% compared to 2018, with a total transaction value of $9.4 billion.

During the pandemic, however, use of this office space changed dramatically. Stay-at-home orders forced many office employees to work from home, while businesses dependent on in-person interactions with customers either reduced capacity or shut down. As a result, by the end of 2020 the value of New York City’s office buildings declined by an estimated 46.1%, according to a recent study.

Current Situation in Office Real Estate

May 2023 marked the first time that office occupancy rates in New York City climbed to 50% of their pre-pandemic level. As of early summer 2023, most NYC workers who were able to work from home still did so, 38% of the time — or about two days per week, on average.

These arrangements, known as “hybrid work schedules,” have become mainstream, with 66% of large companies offering them. Remote work can be a cheaper option for companies, allowing them to accommodate more workers in the same space through the use of open-plan offices and first-come, first-served shared desks (known as “hot desks”), as opposed to traditional assigned desks.

As a result, many companies sub-leased their offices and many more expect to shrink their real estate footprints once their current leases expire. A May 2023 report from the real estate company Colliers documented a 75% rise in available office space since March 2020, with Downtown Manhattan experiencing the sharpest slowdown — one-in-five offices stand empty in the neighborhood that is home to the stock exchange and the World Trade Center.

Projections the Future

A recent study by economists at New York and Columbia Universities found a strong connection between a firm’s remote work policy and its demand for office space. The study estimated that a firm’s need for office space drops by 5-10% for every extra day its staff works from home. For businesses whose employees are expected to be in the office 4-5 days per week, this decline was about 10%. For companies that are fully remote or allow workers to be in the office one day per week, the decline was closer to 25%. The researchers found that, if current trends continue, NYC’s commercial property market could shed 44% of its value by 2029, translating to a $49 billion loss.

The difficulties experienced by the commercial property market are compounded by other factors. The Federal Reserve’s decision to raise rates to combat inflation has hurt borrowers’ ability to finance new deals and service existing debt. In a decision that stunned the market, a real estate investor recently handed the keys to a Downtown Manhattan office building back to its lender without trying to restructure the suddenly unwieldy debt payments.

Implications for Other Areas of the Economy

The office market matters in a city where property taxes account for about 20% of annual tax revenues. A shrinking tax base will potentially affect everyone as further revenue loss may constrain the city’s budget for essential services, such as trash collection, road maintenance and public safety.

While there's been a recent lack of demand for commercial real estate, the residential market has gone into overdrive. A recent paper from the National Bureau of Economic Research suggests that requirements of remote workers—such as space for a desk or office and the extra time at home—have contributed to skyrocketing housing costs.

One solution to the housing challenge is obvious: turn commercial spaces into housing. Converting empty offices into apartments would preserve tax revenue and offer a solution to the city’s housing problem, while also bringing more people back to office districts. Until recently, developers have shown little interest in the idea due to the cost and complexity of such conversions. However, falling office values may force both private developers and city government to collaborate on more adaptive reuse projects in the future.

Where Do We Go From Here?

As NYC treads into the post-pandemic future, it faces a critical decision: adapt to the “new normal” of hybrid work or let the office market languish in hopes of office workers returning at some future date. If the last three years are any guide, people still desire to live in large cities. However, the way in which businesses and governments respond to the emerging importance of “social districts”—neighborhoods dense with housing, restaurants, coffee shops, and co-working spaces—will determine whether NYC will continue to be the powerhouse of global business in the future.

As the noted writer and urban thinker Jane Jacobs wrote in a 1958 article for Fortune magazine: “There is no logic that can be superimposed on the city; people make it, and it is to them, not buildings, that we must fit our plans.”