Of AI, Tortoises, and online safety

The world is split, riven, and – as we so often observe from the three outposts of the Small Data Forum – like never before. Milk or tea in first? Red sauce or brown sauce on a sausage sandwich? And is it still acceptable to say “Happy New Year” after Blue Monday (the third Monday in January and officially the most depressing day of the year, which this year was also your correspondent’s birthday)?

We three podders from Plague Island seem to be in the “Aye” camp for the third of these modern dilemmas, particularly as this – episode 65 – is our first digital emission of 2023. The year in which things can only get better, as Thomas notes, doubtless inspired by the work of D:Ream.

So Happy New Year, podcats!

Whither online safety?

We start our sideways look at the uses and abuses of data big and small by focusing on the UK’s much-delayed Online Safety Bill, likely to finally get Parliamentary time in 2023.

Thomas neatly summarises an excellent recent article from Reuters on the topic headlined “Tech bosses could face jail after UK govt backs down over online harm”. There’s general approval for this move, though the latest Culture Secretary through the door has plans to exempt “those who act in good faith”. Fines levied could be up to 10% of turnover, bringing the bill into GDPR territory.

Neville believes that the Bill will undoubtedly be imperfect – indeed, he describes it as “half-baked” – although he believes it’s probably as good as we’re going to get and certainly better than the oceans of nothing we currently have. Anything that starts with the intention of preventing another tragedy such as that of Molly Russell should be applauded. Social networks SHOULD take responsibility. He points to the plain language and clarity of the ongoing summary of the bill on the Gov.uk website, which identifies what’s in, what’s out, what’s changed, and likely time of passage.

Back to the GDPR future

Sam takes the parallels with GDPR further. The successes of GDPR – a record €2bn in fines levied (if not collected) by different EU member state Information Commissioners in 2022 – also talk to the potential limitations of the UK Government’s Online Safety Bill. They’re likely to be local fines for local subsidiaries of Meta, Alphabet et al., and gaol as the ultimate sanction for local bosses rather than the Zuckster.

Can – Sam wonders – local jurisdictions have power over companies headquartered in California, Dublin, Luxembourg, or wherever they shift their power base to in order to be most tax efficient (aka dodging)? 

Thomas laments the recent move in the Bill to make age restrictions a requirement for self-policing; we’ve seen so many times before how poor Big Tech has been at self-policing. He also believes it’s a shoddy move to blur the distinctions between what is “legal not harmful” and freedom of speech.

This leads to a discussion of the bile spouted by Sun columnist and all-round gammon Jeremy Clarkson about Megan Sussex. While not online safety, there’s not a little Schadenfreude as we catch up in real time with the news that Mr Clarkson has been dropped (or at least suspended) by Amazon Prime from its flagship Grand Tour, lads-n-motors show.

The Daily Fail reports that some neaderthal, incel subscribers are “ripping up their Prime subscriptions in anger”.

Tate vs Greta

We also can’t resist a diversion down the Andrew Tate vs Greta Thunberg twitter spat, though as Neville and Sam agree, the grotesque crimes that the Brothers Grim Tate are more about offline safety than online, albeit mediated via social platforms.

The glory of Greta’s witty, pithy ripostes have been well covered elsewhere, including her excellent observation that his failure to recycle pizza boxes is what finally tipped off Romanian police to his whereabouts.

The slow drift left

We shift gear, and focus on the excellent work of the Financial Times head chief data reporter, John Burn-Murdoch and his recent aggregation of polling data that demonstrates that “Millennials are shattering the oldest rule in politics”.

It is said that controversialist former British PM, Winston Churchill, said: “If a man is not a socialist by the time he is 20, he has no heart. If he is not a conservative by the time he is 40, he has no brain.” Voting intentions and voting actions of 20-40-year-olds both sides of the Atlantic suggests this is no longer the case.

While Gen X, Boomers, and the (soon-to-be) Silent Generation have all lurched from more left-wing voting before their fifties to more right-wing in their 60s-80s in both the UK and the US, the same is definitely not true of the current cohort of 20-40s (see chart).

Voting intent lurches by generation

We are all encouraged by this trend, and although we’re still a maximum of just under two years from an election in Britain, Sam points out that Politico’s poll-of-polls still shows a 22% delta between Labour (47%) and the Conservatives (25%), figures that haven’t shift much since the Liz Truss junta.

Stacked against this, Thomas reflects that the media ecosystem has growing weight on the ultra-right, driving radicalisation and polemic rather than progress. And don’t get him started (again) about the dire imperative to change the electoral system …

All hail the Tortoise

One way for voters to make more informed decisions is – perhaps unsurprisingly – giving them more and better access to information.

We all welcome the development of The Westminster Accounts, a joint initiative from James Harding’s slow news movement, Tortoise Media, and Sky News. This searchable tool enables granular understanding of all donations all MPs have received, how much and from whom since the beginning of the current Parliament in 2019. (Point of information: that’s three tarnished Conservative Prime Ministers ago).

The Westminster Accounts public database

While there’s no smoking gun waiting to be detected – and we’ve all had a good rootle around – Thomas is encouraged that putting this public domain information in the public’s hands can help individual citizens get their heads around the Tufton Street cabals and the murky, murky worlds of All-Party Parliamentary Groups (APPGs). Many of whom – including new PM Rishi Sunak – appear to be getting AI powerplayer ChatGPT to write their speeches for them.

The inevitable ChatGPT chat

There’s a heck of a lot of negativity about ChatGPT right now.

Thomas directs us to a typical naysayer, one Naomi S. Baron, the Professor of Linguistics Emerita at “American University” no less. Her article on The Conversation carries the headline “How ChatGPT robs students of motivation to write and think for themselves” and its content is typical of much of the criticism of OpenAI’s “new artificial intelligence program”.

Sam takes a rather different stance amid all this doom and gloom.

Human progress comes about as a result of successive generations standing on the shoulders of giants from previous generations. We’re recording our podcast on Zoom thanks to mining of metals and other rare earth elements, the discovery of how to master electricity, Marconi, Bell, Turing, Jobs, Gates, Yuan et al. We benefit from previous innovators so we don’t have to do what previous generations did.

Examples of using ChatGPT as an assistant to do some tasks that were previously onerous abound.

Sam cites: Ryan Reynolds’ self-referential use of the platform to create an ad for his Mint Mobile business; a market research boss who routinely uses it to create discussion guides for qual research and then has his teams – who’d have taken a day to do what ChatGPT can do in 30 seconds – to finesse it for his client’s specific needs in but half an hour; and, Key Person of Influence author Daniel Priestley running webinars on using it to write questions for his powerful scorecard marketing platform, ScoreApp. (If you want to know what kind of data storyteller you are, check out your correspondent’s very own Data Storytelling Scorecard, produced – by humans, not AI – using ScoreApp).

Neville recently used ChatGPT to write a series of tweets to promote a local tweet-up, which he then finessed and personalised. He can understand the concern of educators where pupils are using it to write their essays for them, although there are at least two services that aim to detect this and more, AI-driven solutions will doubtless emerge as checks and balances.

He points to a very recent article on CNN about CEOs at Davos openly embracing ChatGPT to write their work emails , citing Jeff Maggioncalda – the CEO of online learning provider, Coursera: “Anybody who doesn’t use this will shortly be at a severe disadvantage. Like, shortly. Like, very soon.”

Even better, Neville asked ChatGPT to create a movie treatment for the third in the series of Knives Out movies, blogged about an excellent plot line the AI platform created, and attracted the interest of a Hollywood producer. He tells the story on one of his blog platforms. Just remember your fellow podders, Neville, and be sure to invite us to the red-carpet première. Move over Rian Johnson …

Towards a new curriculum

Neville does believe that using ChatGPT will lead to job losses, with those creating copy – the lowly account executives in agencies – replaced by OpenAI and others’ platforms. That said, he believes that there will be demand for other skills – like fact and reliability checking.

One thing we all agree on is that, while ChatGPT is a very useful assistant, no-one should ever accept what it produces as being true. Hence the need for human curation and verification. Not least because the current iteration of ChatGPT is only trained on what the world knew up until 2021.

Thomas ends our 45 minutes on a note of optimism about how this might change the nature of education. The future of education should be very much more metacognitive – thinking about thinking; critical faculty – rather than regurgitating facts.

These are topics – ChatGPT, AI generally, and the future of education – that we will embrace further in future episodes.

Listen to Episode 65:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.