Intro
Implement AI Assistants, If we are to believe some of the bleaker headlines about AI, almost no one is safe from destruction by automation. Apparently, the livelihood of translators, scientists, mathematicians, writers and even poets They are already under threat. No wonder 32% of UK employees think technology could make their roles redundant, according to a survey published by the Office for National Statistics in the fourth quarter of 2023.
However, the same research found that 28% of workers believe AI could make their jobs easier. This is the argument of many technologists who, unlike attention-grabbing pessimists, argue that generative AI tools like ChatGPT are much more likely to help workers than replace them.
They argue that digital assistants powered by AI can serve their human masters, rather than exasperate them as Microsoft Office’s much-mocked Clippy feature used to do. They can remove the monotony of people’s daily work, speed up processes or use data in new ways.
And here’s how sectors ranging from retail and logistics to marketing and legal are exploring the potential of AI assistive technology. Influencer marketing agency Billion Dollar Boy, for example, is streamlining its creative process by using Midjourney and ElevenLabs to conceptualize ideas and quickly produce mockups and storyboards, reports its global CMO, Becky Owen.
Meanwhile, international law firm Cleary Gottlieb uses GenAI to scan its databases and create a summary of its lawyers’ relevant experience before meeting with a new client. That document “won’t be ready to ship immediately,” says managing partner Michael Gerstenzang, “but it’s a pretty good first draft.”
Most creatives, at least once in their careers, have experienced paralyzing fear when faced with a blank page to fill, resulting in procrastination and even writer’s block. GenAI-based assistive technology can help overcome these issues, according to Stack Overflow, a coding knowledge hub that recently signed up with Google to power the search giant’s Gemini AI model.
“We think it will be much easier to write code than in the past. When I started coding, I did it by hand and the only reference points were textbooks. For that first draft, GenAI will be able to generate all this critical content,” explains Stack Overflow CEO Prashanth Chandrasekar.
Do you really need AI? implement AI assistants
Companies that implement a new AI tool without carefully evaluating its potential cultural and practical side effects beforehand risk alienating employees or even making their lives more difficult. Therefore, a company must identify whether the technology it is interested in can generate a net benefit or not, and a simple test can be of great help in this regard, according to Peter van der Putten, assistant professor of AI at Leiden University. .
I would advise any company faced with this choice to consider the following questions: “Will this tool automate manual work to sufficient levels of quality? Will it improve the lives of customers and employees? Will it lead to better business results?
If a company can answer all of these questions positively, it should come up with appropriate use cases to experiment with. It would be necessary to create pilot and control groups to measure the impact of the tool, as would be done with any new IT, emphasizes van der Putten, who is also director of the AI laboratory at the American software company Pega.
While it may be tempting to choose a low-value, low-risk use case to test, he recommends choosing an application that allows for “rapid measurement of success” and can also have a big impact once scaled.
Owen points out that such trials naturally involve a certain amount of error. With this in mind, Billion Dollar Boy has established a set of guiding principles designed to ensure that you can benefit from AI in a way that benefits staff and customers. The agency has also created a working group to look at new AI tools and their uses. This group will inform the rest of the business about the latest developments in the field.
“There has been a rise in integrated AI tools, each of which promises efficiencies, but can be clunky and add time to work processes,” Owen says. “The truth is that we may not all immediately arrive at the right solution. The key is to have an open mind.”
Where AI Assist Pilots Can ‘Hit a Wall”
Despite all the excitement about GenAI, there are several pitfalls that companies trying to implement it must avoid.
Chandrasekar recounts a meeting he had with 15 CIOs from the banking sector, all of them interested in realizing the huge productivity gains promised by GenAI. Three months later, these IT bosses “hit a wall” when trust issues related to data privacy and security arose after the pilots.
CIOs were concerned that the data they had been feeding into the tools was “literally going to their competitors’ banks,” Chandrasekar says.
Given what is at stake, ownership becomes a “hot potato,” he adds. “You’re betting your career that this is going to work when you’re pretty early in the hype cycle.”
AI tools must address the credibility issue by adding context, such as quotes, to assure the user that their output has not been poisoned by hallucinations. That’s the opinion of Cassiano Surek, CTO of digital design agency Beyond.
“Given the data-rich nature of AI assistance, ensuring high-quality, relevant information is available will be key to its effective use, as inaccuracies can quickly erode trust,” he says. “AI assistants must be able to cite their sources and have virtually zero hallucinations for such a business-critical use.”
Data privacy guarantees were vital for fashion retailer Asos when it launched an AI-powered code completion tool in September 2023, after a successful pilot earlier in the year. The company had used 90 employees (a group large enough to provide useful feedback on potential issues) to test GitHub Copilot before making it available to all tech staff.
Dylan Morley, principal engineer at Asos, reports that measuring the impact of these types of tools is a topic of “big debate” across the industry. But he adds that this is a more complex issue than simply adopting a tool and expecting, say, a 10% efficiency gain.
“You may instinctively feel that Copilot is faster at solving certain scenarios, but there is a broader question about efficiency in the technology. We are efficient when we move towards strategic objectives and deliver value to customers.”
Morley argues that companies could focus on areas other than adopting AI tools if improving efficiency is truly their primary goal.
“Teams deliver software, so team efficiency is the important thing we care about,” he emphasizes. “Managing time spent in meetings, reducing context switching, improving build and deployment pipelines – all of these things can have a much greater overall impact. “You can be incredibly productive when there is a well-selected set of priorities and a tight feedback loop, and when you know exactly why you are building something and avoid distractions.”
While AI tools clearly have a role to play, it is important to manage expectations about the extent to which they will help.
Morley believes they can save people time they likely would have spent on intense work, but adds a caveat: “What you do with that saved time—how to reinvest it to ensure you see productivity gains—is the important point. .”
Three essential considerations about AI
