“According to recent research, nearly 40% of tasks performed by public-sector workers could be at least partially automated using a combination of AI-enabled hardware and software………... If the UK’s public sector adopts GenAI tools, it could save £17 billion by 2035, but delaying implementation might result in a loss of £150 billion in economic benefit. Additionally, local councils are already using GenAI tools to streamline routine tasks, from meeting minutes to public queries.”
This is a statement directly after enquiring with Copilot, the desktop AI available from Microsoft. This exact number was echoed this month by Tony Blair at a high level presentation. This very same number was featured in the June 20, 2024 Microsoft report “Harnessing the Power of AI for the Public Sector” which quoted Dr Chris Brauer, Director of Innovation at Goldsmiths University.
According to numerous commentaries (Politico, TechRadar and Forbes) Tony Blair’s research was apparently the result of efforts consulting predictions from ChatGPT!
To me, this a perfect example of how AI learning can magnify and lens a single opinion (truthful or not) and create an “alternate fact”. This is the growing danger that we are now seeing across all of society. Social Media ‘false news’ are the roots of the problem, and then this is channelled through AI, which is then the food for more ‘false news’! The end-result is that this misinformation becomes fact for some people, and history and real facts become distorted. There are endless examples today that are too obvious to repeat.
The use of amplification of lies is well established. Propaganda is a modern Latin word, a form of propagare, meaning 'to spread' or 'to propagate'. Propaganda, during the reformation, was fuelled by the spread of the printing press. Pogroms and lynchings also were often the result of false rumours. Into the 20th Century, through the first and second world wars, propaganda was heavily employed as the mass communication methods improved. Fast forward to the early 21st Century where mass communication, via social media has led to false news spreading prolifically to those willing believers.
However, with AI, it could go one step further, and we just may not be able to differentiate the source of the information that we gather at all. Unless humankind manage to build-in safety mechanisms, (such as a compulsory requirement for multiple ‘trustworthy’ and unconnected references per piece of information provided), our reliance on populist information sources in everyday life may diminish, if only for self-preservation. Potentially, some of the very best benefits of AI could be lost.
There is an old proverb: ”If you say something enough times, someone will believe you!” – Never a more truthful word said.
Comments