5 things you need to know about AI this month

Recent developments in artificial intelligence (AI) include the creation of a powerful new natural language processing tool called ChatGPT, developed by OpenAI. The tool uses deep learning to understand and respond to natural language inputs, making it a valuable tool for businesses and organizations looking to improve their operations and productivity.

So far, so good but now it's time for some human intervention, because everything you read above was written by a machine. The ChatGPT AI programme from San Francisco-based Open AI was responding to my prompt asking it to: "Write a news summary that reports the latest developments in artificial intelligence."

ChatGPT's report went on to reference other recent launches of open-access creative AI platforms, spitting out the following paragraph in a matter of seconds:

In other AI news, researchers at the University of California, Berkeley have developed a new machine learning algorithm called DALL-E that is capable of generating original images from text descriptions. The algorithm has been used to create a range of images, from photorealistic to highly stylized, demonstrating its potential for applications in a variety of fields.

The ChatGPT platform is still in trial mode and the company is open about its limitations. For example, the reference to the University of California, Berkeley in the above paragraph appears to be inaccurate, and the launch of DALL-E is somewhat dated in a news context.

OpenAI made the ChatGPT tool available for free public testing on 30 November, according to Reuters. Within a week of it being unveiled, over a million users had tried it, said Sam Altman, co-founder and CEO.

A tool like ChatGPT could be used in real-world applications such as digital marketing, online content creation, answering customer service queries or, as some users have found, even to help debug code, according to Reuters.

Open AI, a non-profit, says its mission is to ensure AI benefits all of humanity. Journalists like me, a naturally sceptical tribe, will be viewing this technology nervously.

The annual survey on the state of AI from McKinsey's QuantumBlack artificial intelligence division shows business adoption of the technology has doubled in the last five years. Half of the survey's respondents said their business has adopted AI in at least one business area. That's up from 20% of respondents in 2017.

McKinsey also found that AI is being embedded into a wider range of business capabilities. The average user of AI in business is now using the technology in 3.8 applications, compared to 1.9 in 2018.

The use of AI in business covers a spectrum of applications, including process automation, digital twins and facial recognition.

The EU rights watchdog says applications that use AI to predict crime and moderate online hate-speech should be free of bias to avoid discrimination. In a report that based its findings on a 2021 study, the EU Agency for Fundamental Rights (FRA) said algorithms based on poor data could harm people's lives.

"We need a system for assessing and mitigating bias before and while using algorithms to protect people from discrimination," said FRA Director Michael O'Flaherty.

EU countries have agreed a common position on draft AI rules ahead of negotiations with EU lawmakers, but drew criticism for not adequately addressing the issue of facial recognition.

The European Commission proposed the AI rules last year, seeking to catch up with China and the United States in a technology used in smartphones, computers, self-driving cars, factories, online shopping and advertising.

European Consumer Organisation BEUC said many important issues have not been addressed, such as facial recognition by companies in public areas, while provisions classifying systems as high risk have been watered down.

Amazon.com is planning to roll out warning cards for software sold by its cloud-computing division, in light of concern that artificially intelligent systems can discriminate against different groups, the company told Reuters.

Akin to lengthy nutrition labels, Amazon's so-called AI Service Cards will be public so its business customers can see the limitations of certain cloud services, such as facial recognition and audio transcription.

If you were intrigued by the story about bias in AI above, take a look at this article on the need for a robust governance framework to ensure the responsible use of AI. It provides a checklist of best practice to guide organizations deploying AI technology.

Science fiction has long focused on the destructive potential of AI. An early example was the intelligent computer HAL in the film 2001: A Space Odyssey. But what if AI was the answer to saving the planet? Have a read of this feature which looks at the role AI can play in helping the world meet the UN's Sustainable Development Goals.

The retail sector is already 40% automated, but this could jump to 60-65% over the next three to four years. Why are more retailers using robots?