AI at our offices and workplaces
When it comes to technology, as we have seen in the past 2-3 decades, it is easier for us to adopt and adapt it in our personal lives, rather than in our offices, and workplaces. It is simpler to change the way we listen to music, read books, order groceries, and entertain ourselves. It is more difficult to do that in the offices, as the use of computers, laptops, Internet, excel spreadsheets, and other software proves. The problem is not just about the employees, Employers and firms need to adjust, and inculcate technology. In addition, tech has inherent issues.
According to an article in the latest issue of The Economist magazine, this is true, and new tech, especially the radical ones like Artificial Intelligence (AI), needs to deal with problems related to employee behaviour, organisational mindset, and technology itself. For example, according to a professor at Wharton School, employees use tech and, hence, they are in the best position to find ways to use it. But, in terms of behaviour, the workers have the best excuses to avoid tech. We take more credit for work done by AI, and are scared to become redundant, or show that they have more free time because of tech tools.
In many cases, the obstacles are in the form of conveniences, and not worries and concerns. The fact is that AI-generated outcomes need to fit into a routine, a habit, if you may say. Employees are used to certain workflows, systems, and procedures in most cases. We are slaves to how we spent the 8-10-12 hours in the offices. Like there are specific times for coffee, cigarette, and gossip breaks, each hour of actual work is counted for every day. If the work of the AI tool does not fit into the regime, we shun it. We discard the results, even if they are beneficial.
According to The Economist, Glowforge, an American firm that makes desktop laser-cutting machines, emailed summaries of daily sales calls to its staff. “Every single sales rep had routed it directly into the bin,” says Dan Shapiro, CEO. “It was too noisy, and it did not have a place in the rhythm of the team,” he adds. Glowforge decided that the recordings were not emailed, but discussed weekly with the salespersons. “You can have a superior product, but if it does not fit into somebody’s work, it is tough to get adoption,” says a head of tech in another firm.
The same happened with computers, desktops, and software. Although they made life easier in offices, people dragged their use for years, even decades in Indian banks, for example. The reasons were the same: attempts to avoid them, be scared and fear tech, and lack of tech’s place in daily routines and habits. This was true of the Internet too. Younger people easily adopted it in a hurry, compared to mid-level and senior ones. The former were flexible in their approach, and the latter two were set in the way they worked, and spent their time.
In technical terms, there are inherent problems with technology. Most tech, like laptops, software, and Internet, have several general issues, and can perform a wide range of work. Even specific software works this way. But firms need tech to perform specific and restricted jobs. Hence, instead of being trained on large-language models, specific AI tools need to be based on the right data, and small-language models. For instance, a tool to deal with logistics, and ordering products, does not need to tell a person about economic theory. It is not a generic search tool, and does need to behave like Google AI, or ChatGPT.
What is crucial, therefore, is that firms need to feed the AI tools with the right and pertinent data. In addition, they must find employees with the skills to handle, and work with the tool. This may require training workshops to equip existing employees, as was the case with computers, and Internet, and now with AI tools. Internal valuations to determine the quality of output, outcomes, and results are critical. Since the outcomes are mostly fuzzy, and subjective, firms rely on humans to define what is good and relevant. Finding, and harnessing this expertise inhouse, or outside, is a challenge.
Tech is not something that impacts all divisions, departments, workflows, and systems the same way. Experts dub it the phenomenon of uneven distribution of AI tools. Since some processes become faster, and efficient, and some do no, the former crowd in the latter. Most tech firms now believe in the mantra, “Demo, not memo.” This means, “You have stopped having the bottleneck at how quickly you can write the code, and now you have got the bottleneck at how quickly you can review the code,” says another head of AI. AI improves invoice processing, and creates panic within the finance team.
The uneven pattern implies that each department does not need to adopt AI in the same way, or at the same level. Firms that believe in letting loose tech across the divisions suffer, as they find, like in one example, that 85 per cent of the value generated by AI tools was attributable to 15 per cent of the applications. So, firms believe in a focused approach through an analysis of where tech will have the maximum impact, and where tech needs to be left out. In essence, some parts of workplaces become nerdy, and some stick to traditions.
Let us apply the same principles to newsrooms since we are part of this newspaper, and readers need to understand why newspapers have changed, both in print and online. The most impact of the Internet and computers are in the post-production stage, sending the data, printing, and even deciding the circulation mode. In online, this implies, uploads, and views. Since tech was deemed crucial at the production stage, newspapers got rid of designers, and even reporters, and favoured multi-taskers, who could edit and use design tools, and content creators, who could steal at will from others, and rewrite them.
So, what we have are badly written content, the same everywhere, sadly edited articles since there is little time, and amateurish designs that are templated, and lack creativity or originality. The result: boring and monotonous magazines and newspapers that look the same. It is akin to seeing the same hairstyles in markets, once a specific one turns viral, becomes addictive, and everyone adopts it.
