Discussion of artificial intelligence often lacks nuance. While proponents highlight the power of AI to provide advances in almost every area of life, there are many who fear AI may prove to do more harm than good. But with AI models like ChatGPT increasingly moving from novelty to business tool, many may find themselves wondering how, rather than if, they can best — and safely — use AI.
TFN recently asked six founders for their views to find out how they are using AI and what they think the future holds.
The different paces of AI adoption
Perhaps unsurprisingly, among a group as diverse as founders, adoption of AI has been mixed as they consider how AI could be used, with some gently testing and considering the potential, while others have leapt in to enjoy the benefits.
Loan investment platform Nectaro, for example, is allowing staff to use ChatGPT to save time. “For us, it serves as an effective tool that enables us to optimise the time we spend on routine, mundane tasks,” CEO Sigita Kotlere told us. “Our employees have observed that since the implementation of ChatGPT and the reduction of effort required for simple yet important tasks, creativity and the speed of strategic decision-making have significantly improved.”
At the other end of the spectrum, Irina Bednova, co-founder and CTO of customer service telephony provider Cordless, says the company has embraced AI, with positive effects. “We use ChatGPT to give our users — customer support teams — access to what is happening in conversations with customers,” she said. The nature of telephone support meant that little data was captured. AI has changed that. “To be able to spot trends in customer queries and proactively react to them has been a dream for customer service managers.”
The risks of AI’s errors
However, startups are also aware of AI’s potential to make mistakes. Examples in demonstrations, such as getting facts about exoplanets wrong, may only be embarrassing. But for businesses, an error could be critical. They are aware of the need to ensure their implementation has safeguards.
Martin Kōiva, the founder and CEO of Klaus, the customer service and quality assurance app, told us they had ”taken great care to prevent LLMs from going rogue.” This has meant both practical and engineering solutions. “We use their generative capabilities only when necessary,” he said, “and we prompt engineering to force the output within controllable bounds and have put guardrails in place for both input and output.”
However, others highlighted that while individual startups will usually take a cautious approach, AI is unregulated and prone to abuse. Ed Thompson, the CTO at Matillion told us, “In the UK, we are still in need of regulation and guardrails for generative AI to prevent some potentially negative outcomes, whether that’s illegitimate data-sharing or use for phishing and defamation.”
The impact of AI on people’s jobs
One of the biggest fears has been the risk to significant numbers of jobs. Some reports have suggested as many as 14 million jobs may disappear in the next five years as AI replaces human workers.
Founders recognise that the job market is likely to change, and several noted that there would be a need for reskilling. However, for most, AI is just another development of the job market, and one that will mostly result in people having better jobs.
Klaus’s Kōiva, for example, said, “our goal isn’t to replace humans with AI but to use AI to make the jobs of specialist teams responsible for conversation quality easier and more efficient.” For him, AI helps increase the capacity for high-quality human roles. “We use ChatGPT to automate routine, repetitive tasks prone to human error. This allows our customer service teams to focus on more complex tasks that require human judgment.”
Cordless’ CTO Bednova sees a future where AI effectively becomes an employee assistant, “we see ChatGPT not as a job-killer, but as a super-powered sidekick for the workforce,” she said. “Far from wiping out jobs, AI like ChatGPT is set to transform them, pushing the boundaries of what we can achieve. It’s about automating the mundane to make room for creativity and the human touch.”
Ally Fekaiki, the founder and CEO of the personalised employee benefits platform Juno, however, did strike a note of caution. “Projections around the number of jobs set to be lost to AI should be taken seriously,” he said. And while he noted the need for regulation, he also said startups have a part to play, “all businesses have a responsibility to use AI ethically, to think carefully about how their use case fits in with the bigger picture, and to keep an open dialogue to ensure we’re working together to get it right.”
The ongoing need for humans
One area that all our founders agreed on was that however powerful AI might be, there would always be a place for humans. Matillion’s Thompson noted that, “people prefer to talk to another person over a machine as humans will always better understand the context of an interaction than machines, and that is unlikely to change. There will always be tasks that require the certain and considered human touch.”
Several also pointed out that, so far, AI is not, technically, creative, nor can it make judgments. In both cases, it can only interpret the data it has, and attempt to present that in response to the prompt. Nectaro’s Kotlere points out that, “AI can indeed add value in areas like data analysis, pattern recognition, predictive analytics or repetitive and mundane tasks.” However, she adds, “there are certain areas that will likely remain the domain of humans, such as creative and innovative outputs, tasks that require a lot of emotional intelligence and empathy, or ethical decision-making.”
Emma Obanye, CEO of OneTech, goes further. AI like ChatGPT, she suggests, can empower people who might otherwise lack the tools or skills to unleash their creativity. “For creatives with an incredible imagination but limited access to high-end tools or those who haven’t been trained technically, it’s truly breaking down structural barriers,” she said.
However, her positivity about human creativity may have carried a note of the pessimism that some have for AI. “True creativity will still be something in the domain of humans,” she said, but that might not last. “Until AI merges with quantum computers, then we’re all in trouble!”
As AI continues to advance, founders recognize its potential while acknowledging the need for responsible implementation. The consensus is that AI should augment human capabilities, not replace them. By understanding the limitations and strengths of AI, businesses can harness its power to enhance efficiency, creativity, and innovation while ensuring a harmonious integration between humans and machines.