Generative AI Trends: Trust & Responsibility

The impact and evolution of Generative AI

The focus on Generative AI definitely peaked, then became a steady drum beat this conference season, as did the approach of businesses to this new technology.

Through the year we saw a shift in focus from firstly exploring the new technology, from implementing and learning from small implementations through to larger scale deployments of some of the more commonly used models.

It was clear throughout the year that Generative AI is undoubtedly the next big revolution in Technology. But questions remain. Can we Trust it? and how to best use it?

Gartner’s Head Analyst for Cloud and AI shared a wrap up of the year AI implementation at the Gartner Symposium and our key take away was a classic one – with great power, comes great responsibility.

For AI to become the partner we need it to be we need to be able to trust it first.

• Forty-five per cent of organisations have increased their AI spend this year largely due to the progress of large language models.
• Nealy half have deployed AI technology for process automation, risk and fraud detection etc.
• Of the forty-eight per cent who have deployed AI technology, nearly three-quarters are running 100 or more models.
• Forty per cent of those that have implemented AI technology had privacy or security breaches which whipped out all the positive gains

What also became clear this year is that prompt engineering is becoming a key skill within organisations. It didn’t exist long ago but has quickly risen to prominence with the advent of ChatGPT and other NLP models.

This is because as with any tool, it is only as useful as the person operating it. As a result, we have seen an increase in the investment to educate staff on how to prompt NLP models to yield the desired result.

We’ve also seen a real surge towards the end of this year in small scale deployments in medium sized organisations to help understand the potential of this technology on their own data sets.