AI is now 'Should we?' not 'Can we?'

Ready access to AI is forcing companies to address ethical questions

Tom Allen
clock • 2 min read
AI is now 'Should we?' not 'Can we?'

AI is so ubiquitous today that the real challenge is not practicalities, but ethics.

2023 has been the year of AI, especially generative AI like ChatGPT. The technology has become so commoditised that it's used in industries from travel to consultancy

"AI is becoming much more pervasive and less limited," said Dan McMahon, head of innovation management at Hymans Robertson, speaking on a panel discussing AI at the IT Leaders Summit this week. 

"Copilot being released into the entirety of the Microsoft 365 suite puts [AI] into many more people's hands." 

Companies' use of AI differs heavily. Hymans Robertston is using it in its core systems. Bank of America, on the other hand, focuses on heavy data analysis, said director Rahul Mundke. 

"We capture pretty much all data coming into the trading platform," which makes it "very difficult" for a person to manage, he said. It's a perfect use case for AI; "finding bottlenecks and new workflows, looking at how to improve the customer experience." 

Trust, transparency and ethics 

Because AI tools are now so easy to access, companies have to ask and answer important ethical questions when implementing the systems. 

"AI needs to explain itself: how it makes a decision, and if that decision is transparent, ethical and trustworthy," said Huseyin Seker, professor of computing sciences at Birmingham City University. "You need to consider how you're developing and deploying it, whether it's rule-based or data-driven, and mostly if you trust it." 

That's especially difficult when you have to buy a system in, rather than building it. 

"We're a Microsoft house, so we're trialling the OpenAI service," said Dan. "It's more challenging than self-building or getting it off the shelf, as you can't just open the lid and see how it works. 

"OpenAI is getting itself into hot water over training data now, so should we use it? If we don't, someone else could beat us to the punch. 

"It's the should we, not the can we."

There are still questions to answer about AI and data. For example, what happens if a user withdraws their consent after their data has already been used to train an LLM? Who owns an AI model trained on public data - the company or the data owners? 

"Regulation and legislation are massive issues and have not caught up [to AI] yet," said Huseyin. "Who is going to regulate this sector? It's a big question." 

And we're still waiting on an answer.

You may also like
Facebook chatbot claims to have a child with 'unique needs and abilities'

Social Networking

Moving fast and breaking things again

clock 19 April 2024 • 3 min read
Stability AI cutting staff in the name of restructuring

Corporate

Following the departure of CEO Emad Mostaque, UK AI unicorn is shedding employees

clock 19 April 2024 • 1 min read
Microsoft injects $1.5 billion into UAE's G42

Artificial Intelligence

Reported 'behind-the-scenes deals' to ensure G42 removed some Chinese tech

clock 18 April 2024 • 2 min read

Sign up to our newsletter

The best news, stories, features and photos from the day in one perfectly formed email.

More on Big Data and Analytics

Even CERN has to queue for GPUs. Here's how they optimise what they have

Even CERN has to queue for GPUs. Here's how they optimise what they have

'There's a tendency to say that all ML workloads need a GPU, but for inference you probably don't need them'

John Leonard
clock 17 April 2024 • 4 min read
Partner Content: Why good data is the foundation of AI success

Partner Content: Why good data is the foundation of AI success

Does your organisation have the right quantity and quality of data to make its AI ambitions a reality?

Arrow
clock 04 April 2024 • 2 min read
Partner Content: Human-in-the-loop - How AI can boost your organisational culture

Partner Content: Human-in-the-loop - How AI can boost your organisational culture

Why it’s vital to consider your organisation’s people when implementing AI

Arrow
clock 26 March 2024 • 2 min read