Business and AI go way beyond than chatbots. Chatgpt, Openai’s revolutionary robot conversationalist, is based on GPT-4, their most advanced artificial intelligence model, which came out about a year ago. Over that span, the market valuation of the US technology sector as a whole has increased by 60%, resulting in $6 trillion in value for shareholders. Revenue growth is beginning to catch up with skyrocketing share values for several tech corporations. The market value of Nvidia, a company that makes chips that powers models like gpt-4, soared to $2trn after the business announced strong fourth-quarter earnings on February 21st. Other digital giants, including Alphabet (Google’s parent company), Amazon, and Microsoft, who are investing heavily in AI research and development, have also seen their share prices rise due to the AI mania.
On the other hand, sales of AI software by big IT companies are still relatively low. In the past year, revenue growth at Azure, Microsoft’s cloud-computing subsidiary, and related services has been driven by factors other than artificial intelligence, which have contributed around 20%. Analysts believe that Alphabet and Amazon’s AI-related sales are lower than Microsoft’s, although the companies themselves do not disclose this information.
These companies will eventually have to start making a ton of money from customers if the artificial intelligence stock market bubble is going to last. All kinds of companies, from financial institutions and consulting firms to movie studios, need to implement Chatgpt-like systems immediately. There has been a cautious approach from businesses towards the practical use of such “generative” AI. This is still a modest step, but it does show that white-collar jobs are evolving with this business and AI collaboration.
What people perform on the job has been radically altered by earlier technical developments. “With the aid of this little machine, an operator can accomplish more correspondence in a day than half a dozen clerks can with the pen and do better work,” stated an observer in 1888, describing how the typewriter’s widespread use led to several employees losing their jobs. The advent of personal computers at the turn of the century later increased productivity among highly trained workers while simultaneously doing away with certain low-level administrative duties. An article claims that computers account for more than half of the change in demand for workers with bachelor’s degrees between the 1970s and the 1990s. The recent surge of remote work, made possible by video conferencing and encouraged by the COVID-19 pandemic, has already altered the routines of white-collar professionals.
Is it possible that generative AI might cause changes of this magnitude? The economy as a whole takes a long time to reap the benefits of technical advances, as was the case with earlier ones. Adapting to new methods of operation takes time for the typical employee at a typical company. Gains in productivity due to personal computers did not materialize for at least ten years following their widespread availability. There has been no discernible increase in overall economic productivity as a result of AI-induced factors. Almost all executives surveyed by the Boston Consulting Group (BCG) believe that “moving beyond the hype” surrounding artificial intelligence will require at least two years. Adoption of AI “has not necessarily translated into higher levels of productivity—yet,” according to recent research by Oliver Wyman, another consultancy.
It is not surprising. No matter how much fun individual employees have with programs like Chatgpt, Google Gemini, Microsoft Copilot, or any of the others, most companies still don’t use them extensively. The US Census Bureau investigates thousands of companies every two weeks if they employ AI in any way. It encompasses both the more recent generative kind and the more traditional sort that businesses used for a variety of purposes before 2023, such as enhancing search engine results and predicting inventory requirements. Fewer than five percent of American businesses reported using AI in February. The figure shows that another 7% of businesses intend to implement it in the next six months. The data also doesn’t show how different different industries are; for example, just 3% of manufacturer and 5% of healthcare companies claim to use it to create products, but 17% of information industry (that includes media and technology) enterprises do.
In September 2023, when the Census Bureau initiated inquiries regarding artificial intelligence (AI), it was observed that small firms were more inclined to adopt the technology than large ones. This disparity may have been attributed to the fact that smaller firms required fewer form-checkings, which facilitated adoption for minnows. Nowadays, AI is mainly used by large firms (those with over 250 people). These companies have the financial means to hire AI teams and invest in what is needed. Morgan Stanley, a financial institution, surveyed major companies and discovered that the percentage involved in AI pilot projects increased from 9% to 23% from the beginning to the end of 2023.
Some multinational corporations are racing to test new ideas to determine which ones are successful. Statistics from the job-search website Indeed show that they are employing artificial intelligence specialists by the thousand (see chart). “More than 300 AI use cases are in production today,” was mentioned by Jamie Dimon, CEO of JPMorgan Chase, last year. The consulting firm Capgemini has announced its intention to “use Google Cloud’s generative AI to develop a rich library of more than 500 industry use cases.” The large German chemical manufacturer Bayer asserts that it has over 700 applications for generative AI.
There are three main types of this “use-case sprawl,” as one consultant puts it: window-dressing, tools for low- to middling-skilled workers, and tools for a firm’s most valued people. The most prevalent of these is window-dressing. The University of Toronto’s Kristina McElheran claims that many companies are renaming their ordinary digitization initiatives as “gen ai programmes” to make them appear more complex. The restaurant technology provider Presto has launched a Gen-AI assistant that can accept orders at the drive-thru. However, a human assistance is necessary for exactly 70% of these requests. One artificial intelligence (AI) disc jockey that Spotify has released plays tracks and makes random comments. After artificial intelligence (AI) showed customers unappetizing pictures of vendors’ food, supermarket delivery service Instacart recently disabled a function that created such pictures. Even large tech companies are integrating their own artificial intelligence innovations into products aimed at the general public. Rufus, Amazon’s new AI-powered shopping helper, is something no customer has ever requested. Google Maps is now more “immersive”, whatever that means, thanks to the use of artificial intelligence.
There may be a greater need for tools designed for less trained laborers right now. Commercially available AI has a few basic uses, such as in customer support. Chatbots may be easily trained to handle most client inquiries because these questions tend to be straightforward and cover a limited number of themes. You could find that some of these efforts are yielding results already. Amdocs develops software that telecommunications firms can use to handle customer support and billing. According to the business, an over 50% reduction in customer call processing time has been achieved through the usage of generative AI. Sprinklr, a company that sells comparable goods, recently said that a client in the luxury goods industry “has seen a 25% improvement” in customer service ratings.
Regular administrative duties also appear to be prime candidates for disruption by AI. Jobs like “easily getting data from Excel files” and “creating a first draft in Word” are among the “top examples” of Bayer’s 700 use cases, which include tedious tasks. Generative AI is being used by certain companies as a smarter search engine. Nasdaq is a financial services company that uses it to aid their financial crime investigators in their quest to identify questionable bank transactions. The business claims that this reduces a process that normally takes 30–60 minutes to only three.
In spite of the complexity of their needs, few companies have begun to provide artificial intelligence tools to their most valuable employees. However, its visibility is also on the rise. Among the first to embrace it were legal professionals. For tasks like as due diligence and contract analysis, the attorneys at the large legal firm Allen & Overy have developed a system in collaboration with the artificial intelligence startup Harvey. The research process at investment banks is being automated with the help of artificial intelligence. An artificial intelligence system at Bank of New York Mellon processes data for analysts each night and provides them with a preliminary draft for morning work. According to the bank, this leads them to “get up at six” instead of “get up at four” in the morning. Small favours. Sanofi is a French pharmaceutical company that employs an AI app to give its executives access to data about many parts of the business in real-time.
A number of businesses are putting this tech to use in developing software. 1.3 million people have subscribed to GitHub Copilot, an artificial intelligence tool for developing code. Google and Amazon both have competing goods. Reportedly, Apple is developing one. Operating firms of the IT giant Fortive “are seeing a greater-than-20% acceleration in software-development time through the use of gen ai,” according to the business. According to Chirantan Desai, COO of business software giant ServiceNow, GitHub Copilot is responsible for “single-digit productivity gains” for the developers at his organization. An Indian business called Konnectify increased their app release frequency from four to seven per month with the use of AI tools. According to Microsoft surveys, once users start utilizing Copilot, they rarely desire to remove it.
One social media platform claims that generative AI has increased the relevancy of search results for users by 10%. This is Pinterest. Bill Ready, the company’s boss, boasted on a recent earnings call that the new models were a hundred times larger than those that his firm had previously used. Investors are taking notice of L’Oréal, a global leader in the cosmetics industry, as it upgrades Betiq, an internal tool for tracking and enhancing the effectiveness of the company’s marketing campaigns. “Some of our brands that have deployed it” have reportedly seen “productivity increases of up to 10-15%” thanks to generative AI, according to L’Oréal.
This does not imply that these firms will require a reduction of ten to fifteen percent in workforce. The concern about a potential AI jobs catastrophe appears to be untrue, similar to previous technology revolutions. It seems that technology is generating more job opportunities than it is displacing at this point. Evercore ISI, a bank, released a survey in November showing that only 12% of companies agreed that generative AI had already replaced human workers or would do so within the next year. While some tech companies have stated that AI is causing them to freeze hiring or reduce workers, there is no indication that layoffs are increasing in the developed world.
Alternative white-collar occupations are also emerging as a result of generative AI. Firms such as KPMG, a consulting firm, and Nestlé, a coffee-to-cat food conglomerate, are recruiting “prompt engineers” who contain the ability to extract valuable responses from artificial intelligence chatbots. “Explainability engineers” are employed by one insurance corporation to assist in comprehending the results produced by AI systems. A consumer products company has appointed a “sales-bot manager” to oversee the use of generative AI by the sales staff.
Even while these changes won’t show up in productivity numbers for some time, they are already having an impact on white-collar professionals’ daily tasks. Sure, there are some positive impacts. With the use of AI, companies can now bring all their previously dispersed internal data, such as meeting minutes and performance reports, into one central location. A consultant, Randy Bean, surveyed companies and found that many of them had much easier success developing an internal “data and analytics culture” than they had before.
There can be some unexpected outcomes of business and AI collaboration as well. A survey for software business GitClear revealed that, over the past year or two, the quality of software engineering work has decreased, even while AI code-writing tools are making software engineers’ tasks easier. After utilizing AI to generate a first draft, programmers may find that it is either too wordy or full of errors. This means they may have more time to study and change code rather than write it from scratch. The amount of work done in today’s offices may increase if other businesses go through the same situation, with artificial intelligence producing more emails and memos, even though these documents are becoming less effective for actually getting things done.
According to surveys conducted by IT company IBM, many businesses and AI is not in books for alot of businesses due to a lack of in-house knowledge in the field. Some people are concerned that their data is overly complicated and stored in separate systems to be integrated. Roughly 25% of American employers outright forbid the usage of generative AI in the workplace. They may be hesitant because they are concerned about the security of their company’s data. Concerns regarding AI-related dangers, such as the potential theft of intellectual property by AI model-makers, have been raised by private equity firm Blackstone and pharmaceutical company Eli Lilly in their respective annual reports. Orange CEO Marie-Hélène Briens Ware revealed last year that the telecommunications operator had implemented data guardrails prior to beginning a trial with Microsoft’s Copilot.
Business and AI is becoming one at some places. As a whole, generative AI still has room to grow before more companies view it as black and white. Microsoft introduced a Copilot in November for its Office productivity suite. Even for those who are well-versed with Office, some early adopters find it surprisingly clumsy and crashy. For as long as the models do not fabricate information, many managers are still hesitant to employ generative AI for more complicated tasks. Air Canada’s artificial intelligence chatbot recently got the business in trouble for providing a customer with false information regarding the airline’s refund policy. Even while that was humiliating for the carrier, it’s easy to think of worse things that could happen. While the marriage of business and AI have a rocky start. Still, everything started somewhere, even the typewriter.
