AI: The Risks and Opportunities

Artificial Intelligence is a hot topic in 2023. As Artificial Intelligence gradually becomes more prominent around the world, we are seeing more and more industry experts speak out on the issue. While some applaud AI and welcome the opportunities that it provides, there are others that fear it, and believe that the risk it poses outweighs the benefits:

“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
Statement on AI Risk from the Centre for AI Safety.

Meanwhile, other experts have taken a different approach:

“…the most common reaction by AI researchers to these prophecies of doom is face palming.”
Professor Yann LeCun (NYU Professor, Meta employee, and considered one of the ‘godfathers’ of AI).

It’s safe to say we’re far from reaching a consensus on whether Artificial Intelligence (AI) is a good thing or not. As a tech company that both caters to and comprises a range of experts, we thought we’d ask around to see how different people saw the threats and opportunities of AI. Here’s what we found out:

Generative AI Enhances Productivity but is not a Substitute for Thinking

“Generative AI (GAI) offers significant productivity opportunities, now accessible to all web users regardless of their expertise. It can assist with various tasks such as coding, blog writing, creating statements of work, and generating digital assets like dashboards, slides, animations, images, and 3D models.

While AI/Machine Learning (ML) technology is not entirely new, its applications have become more prevalent, like product recommendations from large retailers and search engines. GAI can help bridge the skills and knowledge gap, enabling faster learning and easier task execution.

However, a potential threat lies in over-reliance, leading to complacency and reduced problem-solving abilities over time. Those who master GAI will experience significant productivity gains, while those who neglect essential skills may face a productivity disparity.”

Piers Batchelor, Senior Product Manager

When talking to people about how they use GAI for work, a recurring theme is that it’s a useful tool for idea generation or producing a structure. However, in terms of producing an end product, it is not a substitute for the work they do.

This is especially true in the context of AI ‘hallucinations’, where AI produces outputs which are not based on factual information or real-world knowledge. Human input is therefore necessary not only in curating the correct prompts but as a sense check on the outputs.

 

Regulation is Needed to Protect from AI risks

“There are clearly many positive use cases for AI in our daily lives, from optimizing business strategy with predictive modeling, to taking care of mundane day-to-day tasks that could easily be automated.

However, there is also huge scope for entities and individuals to exploit AI for monetary gain, as well as more nefarious purposes like deepfake identity theft, to name just a few examples. It is essential that comprehensive regulations are passed to help tackle these instances of misuse, especially as this technology becomes more widely adopted.

Left unchecked, AI has the potential to be very dangerous to our security and wellbeing as individuals and as a wider society.”

Lizzy Hillier,  Graphic Designer

Government regulation is a prevalent topic surrounding AI. Perhaps most notably, the leaders of OpenAI called for regulation of superintelligent AI systems. The EU is the first to take action in this regard, developing an AI Act which seeks to ensure AI systems are “safe, transparent, traceable, non-discriminatory and environmentally friendly”.

The headline is that the legislation would require companies to disclose what data the AI models have been trained on. Companies would also be potentially liable for what their AI’s are used for. There’s been substantial pushback on this, some criticizing the attempt to regulate such a nascent technology, with US tech executives viewing it as EU protectionism.

The vast majority of people agree that robots and AI require careful management. We don’t need to wait for governments to legislate appropriate AI use. At the organizational level, we can establish clear objectives and guidelines for directing ethical AI use, focusing specifically on transparency, fairness, privacy, and bias mitigation.

The Potential for Data Analytics is Huge, but so is the Risk of Misinformation

“AI, in my opinion, is an exciting and constantly evolving technology, poised to be more widely used in the future. It opens the door to immense opportunities, particularly in data analytics, where it can sift through enormous amounts of information, offer predictive analytics, and provide real-time insights.

However, there are always risks associated with new technology. A major concern for me is the potential for misinformation to appear credible. Even before AI became widely used by the average person, misinformation spread rapidly.

With the ability to create AI-generated images and videos that appear authentic and can depict political figures and world leaders, I worry that individuals will either believe these AI-created images and videos are real, or become skeptical of all news, reaching a point where we can no longer discern what is real and what isn’t.

This could have significant implications for elections and world politics.”

Stephanie French, Data Visualization Developer

Astrato is already employing AI to produce better data insights quickly and more accessibly to business users. Via the intuitive interface, users can generate visualized and text-based BI that can then be shared directly from the platform in just one click.

The implications of AI for democracy is also an important issue. Stephanie highlights the danger of deep fakes in undermining trust in political institutions. Professor Dame Wendy Hall, a computer scientist, echoed this ahead of important elections in both the US and UK in 2024. Cybersecurity is another concern, though there is also significant potential for AI to improve digital security.

We Should Keep Adopting AI but Help Those Whose Jobs are Replaced

“One of the great opportunities of AI – improved efficiency – is also one of the threats. While AI creates new opportunities for skilled roles, it is not always clear that the new job market will balance out lost jobs. In many cases, AI can create far fewer jobs than it displaces, which presents a potential issue for how to address job loss in a rapidly changing work environment.

Even with these concerns, AI is here to stay. It is necessary to balance the undeniable potential for innovation with concerns about job displacement and other ethical issues. Ensuring there are laws and policies protecting people from the negative consequences of AI is key for a sustainable future.”

Luke Tolchard, Product Manager

Fear over job losses has accompanied almost every significant technological development, from the Loom to agricultural machinery and robotics. So far, these advancements, while replacing jobs in one area, have produced new opportunities in another.

The power and broad applications of AI make the concern over job losses especially acute. Siphoning some of the value created by AI towards educating those whose jobs have been replaced is one way of mitigating the damage caused by job losses. This could be for more technical roles in their industry or for entering a new industry entirely.

In Design, There’s Still no Substitute for Human Creativity

“There is a lot of concern (and scare-mongering) over loss of jobs surrounding AI, but I think there are many benefits for designers. AI can save time working on repetitive tasks, such as basic image editing (removing backgrounds, skin retouching) as well as creating color schemes and mock-ups.

It allows designers to speed up their workflow and spend more time on larger creative projects. AI is developing and although it can create impressive graphics and video effects in seconds, it still requires human input, emotions and creativity.

There are some shockingly bad examples of AI not being able to generate human hands and feet.” 
Dan Liburd,Senior Graphic Designer

Writing the right prompts is essential for getting the most out of GAI. At present people with specialist knowledge for the task at hand are the best to do this. Human creativity is essential as it enables tailored content for specific contexts and audiences. This is true not just for graphic design but all applications of GAI.

Ultimately, well-crafted prompts, coupled with human creativity, maximize the benefits of AI while ensuring responsible and meaningful interactions with artificial intelligence.

AI Can Do a Lot of Good, But We Should Be Concerned About Privacy and Bias

I think the apocalyptic, ‘AI is here to kill everyone’ narratives distract from the problems of privacy and bias.
Privacy in that the pervasive datafication of everyday life – ‘smart’ technology and the ‘internet of things’ – creates highly intimate information about our everyday lives, potentially accessible by both unelected big tech companies and increasingly authoritarian governments.

Bias in that perceiving AI outputs as objective truth ignores the biases of the people who built the systems and the data they are trained on.

That being said, there is a lot about AI I am incredibly excited about, innovations in medicine and clean energy in particular. Disruptive technologies are rarely purely good or bad.

James Hurren, Content Writer

The idea that the more sci-fi-esque dangers of AI distract from the immediate threats is common and echoed by Princeton University computer scientist Arvind Narayanan in this BBC article.

Elizabeth Renieris, a Senior Research Associate at the Institute for Ethics in AI at Oxford University, supports the idea that advancements in AI “will magnify the scale of automated decision-making that is biased, discriminatory, exclusionary or otherwise unfair while also being inscrutable and incontestable”. However, action on the more immediate threats of AI doesn’t have to preclude action on the more existential issues.

AI: A Risk or an Opportunity? 

As developers, analysts, and designers, the Astrato team sees AI from a range of perspectives. Despite this, everyone acknowledged both the exciting opportunities presented by AI and some of the potential threats.

The power and prevalence of AI will means it will no doubt create immense value and potential harm. Real as some of the threats may be, the prospects for enhanced BI are equally valid; Astrato continues to stay at the forefront of AI developments and its applications for BI analytics and visualization.