In a few short years, the conversation around AI has changed significantly. What was once a niche technology has gone fully mainstream, both in business and amongst the general public. Three quarters of knowledge workers are already using AI, while ChatGPT has over 180 million users worldwide.
Such widespread familiarity has shifted concerns from the theoretical to the practical. Rather than debating the big-picture implications of an AI revolution, many are focusing on the specific ways that it could change the world for good or bad. With the computational power used to train AI models doubling every six months, it’s no longer a question of whether AI will transform our lives- it’s a question of how.
At Daemon, we’re just as preoccupied with this question as everyone else. We’re hugely excited by the possibilities of AI, but that doesn’t mean we’re complacent about the risks. We understand that a robust ethical framework is needed to ensure that AI benefits everyone, not just a privileged few. We sat down with our resident AI guru Damien Duff to talk about the best and worst case scenarios, and what we can do to move the needle closer to the former.
Connect with Damien on LinkedIn
AI to the rescue
We’re tech optimists by nature, so let’s begin with the positives. Of all the possible applications of AI, perhaps the most exciting is its use as a weapon in the fight against climate change.
Start-ups and nonprofits are already using the technology to track the melting of icebergs, map global deforestation and create detailed maps of ocean floor pollution. This data is far more accurate than anything gathered by humans, allowing more efficient solutions to be designed.
AI can also help to streamline existing environmental initiatives. California-based Eugenie.ai has developed an AI-powered emissions tracking platform which can help companies cut their emissions by up to 30%. Meanwhile, London startup Greyparrot has built an AI system for recycling facilities which identifies recyclable items and prevents them being sent to landfill.
As AI grows more sophisticated, so too will its contributions to the ecological cause. With big hitters like Google DeepMind on the case, the coming years could see some truly staggering breakthroughs.
Healthier humans, healthier society
Moving from the planet to its inhabitants, healthcare is another area in which there is great cause for optimism. 94% of healthcare businesses are already using AI in some capacity, and the AI healthcare market is expected to grow from $11 billion in 2021 to $188 billion in 2030.
Of course, more money doesn’t guarantee better outcomes, but the early signs are encouraging. It’s estimated that AI could reduce errors amongst healthcare professionals by 86%, saving as many as 250,000 lives a year. AI can also be a valuable diagnostic tool. A recent study published in The Lancet Oncology journal found that AI could detect 20% more breast cancer cases than human doctors working alone.
The good news doesn’t end here. As Damien explains, AI has plenty more tricks up its sleeve when it comes to keeping us healthy:
“We’re making leaps and bounds in healthcare. We’ve got better and better models doing things like prediction of protein structure. Understanding how medicines and vaccines will relate to diseases and cells is a massive simulation and modelling problem, but AI is making things a lot faster - amenable. We can now train neural networks to understand how biological mechanisms work. We can simulate human biology far more accurately, allowing us, for example, to see how a vaccine would interact with a virus within the body.”
As well as making us healthier as individuals, AI may be able to fix some of the ills of our society. In a world where politics is driven by emotion and tribal allegiance, AI could help voters sort fact from fiction. This could lead to a better-informed society, a higher level of debate and, ultimately, a healthier democratic process.
An AI-augmented society may also see a rise in living standards. 40% of working hours could soon be automated, giving people more time to focus on social activities and hobbies. As well as making life more pleasurable, this could also have a dramatic effect on inequality. As Damien explains, it wouldn’t be the first time that labour saving devices have acted as catalysts for social change:
“One of the key positives of technology is the agency that comes from putting power in the hands of individuals. The car and the internet did this, but the washing machine is the best example. It took hours of drudgery out of people’s lives, giving them back their agency in the process. AI can do the same.”
Read more about our AI services
Getting down to business
These are inspiring ideas, but business owners will be more concerned with practicalities. In a recent survey by Forbes, 600 business owners were asked about their current and future plans for AI. The results indicated a widespread sense of optimism around the technology:
- 56% of business owners are already using AI for customer service purposes
- 97% believe that ChatGPT will help their business
- 64% believe that AI will improve customer relationships
This is just the beginning. While customer service is a natural use for chatbots, the business applications of AI go much further. As Damien explains, the possibilities are practically endless:
“There are a hundred million things you can do. Whether it’s selling your products and services, keeping your catalogues and institutional knowledge up to date and consistent or keeping track of safety recommendations and compliance, you can automate the gathering of information and the servicing of customers to provide a much better customer experience. You can also reduce the load on logistics, making things far more efficient across the business stack. AI when done right, is like 100 people working to optimise your processes.”
The other side of the coin
.jpg?width=378&height=252&name=pexels-cottonbro-5473955%20(1).jpg)
There are plenty of reasons to be optimistic, but it would be naive to think that the AI revolution will be entirely smooth sailing. For every potential benefit, there’s a potential risk that’s just as powerful.
Take the quality of life argument. AI has the potential to raise living standards, but there’s no guarantee that these improvements will be evenly distributed. AI technology is largely controlled by wealthy individuals, and there’s a real risk that its benefits will be confined to the same privileged few. A recent report by the International Monetary Fund came to a bleak conclusion:
“In most scenarios, AI will likely worsen overall inequality”.
The environment is another area in which AI could be a mixed blessing. It may be a powerful tool in the fight against climate change, but the technology is far from green. The carbon emissions that come from training an AI model back in 2019 was five times greater than those of a car, and there are also grave concerns about the amount of e-waste generated by AI technology.
Worse still, these dangers can feed into one another. Here, Damien describes a scenario that could encompass both the social and environmental risks of AI:
“If we somehow manage to solve the vast problems with inequality and concentration of power, as well as its possible use for violence, and imagine a future with a completely democratised AI, we can then imagine scenarios where absolutely everyone has a great power to influence the world around them, as far as building their own houses, but going further we can imagine people building structures on a whim, consuming natural resources, arbitrary structures strewn across the landscape - we have seen similar issues for example with the relative accessibility of the internal combustion engine and its contingent effects on the environment. Managing this will be a difficult social issue.”
The idea that AI could bring about a healthier version of democracy is also up for debate. Many fear that the opposite could be true, with deep fakes undermining the notion of truth so thoroughly that we completely disengage from the democratic process. Even at their current, primitive stage, political deep fakes have been shown to fool almost half of voters. As the technology improves, this could represent a critical threat. As Damien points out, a breakdown in trust towards our leaders and institutions could be mirrored in a breakdown of trust between individuals:
“We could see a world defined by a lack of trust. A world in which human connection and empathy is disrupted. A world where relationships are mediated through the internet or through AI agents, changing what it means to be empathetic. This could lead to personal breakdowns and social breakdowns across the world. We don’t know. Nobody knows.”
Finally, there are more existential risks to be considered. AI has proved itself adept at mimicking human creativity, with AI-generated songs, books and poems already commonplace. This redefining of artistic expression could fundamentally alter our idea of what it means to be human, leaving us with a profound sense of confusion and loss. Damien has thought about this a lot:
“AIs are making music and writing poems- the sorts of things that we value people doing the most. This could displace our previous selves. It's going to change what is meaningful to us- what it means to be people.”
How Daemon built an AI-powered historical chat engine
Businesses need to be vigilant
Just as there are business-specific benefits of AI, there are also business-specific risks. One of the most concerning is the dehumanising effect that AI could have on customers and employees. As AI gets better at predicting human behaviour, there will be more and more opportunities for exploitation. As Damien explains, resisting this temptation will be a major challenge for businesses:
“You don’t want to treat people as an instrument. You want to treat your customers and workers as human beings. In a world where we have often become subservient to technology-driven systems, there’s a real risk that autonomy could suffer more with more advanced technology.”
AI can also perpetuate inequality in business practices. Machine learning models can pick up and amplify existing human biases, leading to outcomes such as racial discrimination in credit scores.
Finally, there are the security risks to consider. AI-driven cyber attacks could lead to massive data breaches at a time when companies are holding more customer information than ever. AI is a godsend for hackers, allowing less experienced criminals to augment their skills and do serious damage. A recent report by the UK government’s National Cyber Security Centre concluded that:
“Artificial intelligence (AI) will almost certainly increase the volume and heighten the impact of cyber attacks over the next two years.”
As well as being used for hacking, there's also the risk that AI models themselves could be hacked. These models contain huge amounts of data, making them a tempting target for criminals. As Damien makes clear, businesses will need to tread very carefully:
“We’ve seen machine learning models that have been trained up on data from a variety of sources, and then people have been able to get into them and get the data. Models can also be used for negative behaviour such as hacking other businesses. Businesses face risks from AI whether they use AI or not.”
Daemon Principal Cloud Consultant Marc Peiser on the biggest cybersecurity threats facing businesses
Everyone has a part to play
The future of AI won’t be determined by any one individual, company or government. It will be the cumulative result of thousands of decisions, each adding another brushstroke to the final picture. As Damien tells us, this is a responsibility that we should all take seriously:
“I would like for us as a world to continue to pursue this technology because of the benefits. But, at the same time, we need to not only be aware of the negative stuff, but be actively working to prevent it. Every business needs to consider their ethical stance. How am I contributing to the negatives? Am I treating my customers and workers as people? Is this increasing or decreasing agency? What are the environmental impacts of my decisions?”
So what does this mean for Daemon? Like everyone else, we have a stake in the future of AI and want to do our bit to move things in the right direction. We might not have the power to determine the ultimate course of AI technology, but we can strive to make a positive contribution. Here’s Damien again:
“We’re trying to work with businesses that are using AI to drive positive change. We’ve got clients who are streamlining healthcare, offering educational experiences and actively working to make AI safer. We work with our clients and our people to ensure that we're doing the highest quality work, not putting people's data at risk, and steering our solutions towards increasing agency and reducing negative impact. We also have an environmental push to get a handle on AI tools to help with ESG reporting, which we are applying inward first.”
Every transformative technology has its problems, but we trust that the good will outweigh the bad. If you have any questions about AI, ML or any other aspect of digital transformation, don’t hesitate to give us a call. Here’s to the future!