AI - When will they learn? When will they ever learn?!

 

AI was supposed to make the UK benefits system more efficient—instead it's brought bias and hunger A freedom of information request has revealed that an AI system used by the UK government for assessing benefits cases is apparently getting it wrong by a "statistically significant" amount. The admission to journalists at the Guardian emerged after a fairness analysis into universal credit claimants in February 2024. It confirmed that the very tools intended to ensure equity and efficiency may in fact be discriminating against marginalized communities.

Local governments are using AI without clear rules or policies, and the public has no idea

In 2017, the city of Rotterdam in the Netherlands deployed an artificial intelligence (AI) system to determine how likely welfare recipients were to commit fraud. After analyzing the data, the system developed biases: it flagged as "high risk" people who identified as female, young, with kids, and of low proficiency in the Dutch language....

Umm, here's one reason; there are many others!

New technique reduces bias in AI models while preserving or improving accuracy

Machine-learning models can fail when they try to make predictions for individuals who were underrepresented in the datasets they were trained on.

Artificial Intelligence - the Way Forward    The dangers of AI in both education and government services are being documented everyday. The w... more 

They are not alone!

More than half of UK government nutrition advisors are paid by food companies, research reveals

At least 11 of the 17 members of the Scientific Advisory Committee on Nutrition (SACN) have conflicts of interest with the likes of Nestle, sugar manufacturer Tate and Lyle, and the world's largest ice cream producer, Unilever, reports Sophie Borland in The BMJ.

 Training AI programs, known in the industry as large language models (LLMs), involves scraping vast quantities of text or images from the internet.

This information is broken into trillions of tiny machine-readable chunks, known as tokens.

When asked a question, a program like ChatGPT selects and assembles tokens in a way that its training data tells it is the most likely sequence to fit with the query.
https://techxplore.com/news/2024-08-inbred-gibberish-mad-ai.html?utm_source=nwletter&utm_medium=email&utm_campaign=daily-nwletter

This sounds remarkably like the behaviour of some quiz masters, who make an enquiry on a search engine, and look no further than the listed results, which include the terms used in the query, but don't necessarily confirm the answer.

Another example is the lions at Battle Abbey query, quoted in  https://www.academia.edu/116629864/Artificial_Intelligence_the_Way_Forward

And I am sure there are thousands, if not millions more.

There's still hope - common sense prevails

 

Google pulls AI ad that irked some Olympics viewers

Google confirmed Friday that it had pulled an ad for its Gemini artificial intelligence after viewers pointed out that encouraging  your child to use AI to write a fan letter is maybe not the best way to educate them.  May open a can of worms.......

When it's good to be a NIMBY

 

Riots in the UK: Online propagandists know how to work their audiences—this is what we are missing

The frightening scenes of far-right extremists have many wondering how to stop the spread of the propaganda that encourages racism, violence and misogyny.

It's imperative an answer is found; anti-social media is the curse of our times, as some of us predicted, but weren't listened to, when it was still early enough to put restrictions in place.  What is the answer now?

Now there's a surprisde!

 

AI search tools and chatbots may make news less visible and reliable

Evidence is mounting that the new generative AI internet search tools provided by OpenAI, Google and Microsoft can increase the risk of returning false, misleading or partially correct information.

See - https://www.academia.edu/116629864/Artificial_Intelligence_the_Way_Forward

 

Study of key characteristics of UK politicians reveals 'ambition, narcissism, genuine idealism' among common traits

In a new study of politicians' personalities, humor, charm and raw courage are listed among the most important character traits for successful leaders.

How about maturity, experience of the wider world of work, less PPE, more intelligence......