Experts tend to be less certain about their expertise than relative beginners. The Dunning-Kruger effect occurs when someone has a limited understanding of a subject, yet feels overly confident. It reflects the proverb "a little knowledge is a dangerous thing". As people acquire more knowledge about a subject, their confidence in their understanding diminishes.
The Dunning Kruger diagram shows a massive peak in confidence not long after we learn about something, which some wags refer to as the “peak of mount stupid”. It’s the new coder who starts advising veterans on the possibilities in new frameworks, the second jobber who starts telling the boss about why TikTok will change the world, not realising the boss has lived through several world-changing media moments.
The current surge in AI means many of us are at the peak of Mount Stupid, or nervously realising that we were just there a moment ago. But more important than realising where you are on your learning journey into this new landscape is realising that there may not be just one mountain.
Technical experts may assert confident views about how AI will be used by creatives and teachers – but they don’t have the domain knowledge of people working in those fields. An eminent author recently wrote about his fascination with AI, but saw ChatGPT and the current generation of tools as being good only for repurposing content and therefore useful for email spam and content farms (low grade content production to affect search engine results). He may not be at the top of Mount Stupid when it comes to writing for a living with AI, but he is blind to the work that many people do in research, teaching and analysis, where being able to reformat and repurpose content – say from a 300-page company annual report to a summary that can be read by manager or a student – would be an absolute gift.
The more confident someone’s opinions about the future of AI the more skeptical we should be. What are they seeing? What are they missing? Do they understand where they are in a vast range of Mount Stupids?
Don’t bank on it
Another example of no one, even experts, being able to see what’s happening in a complex system…
March 2023: It’s three years since the pandemic really kicked off in the UK.
Here’s one after-effect no one saw coming. When money rushed into tech at the start of the pandemic – the Great Digital Leap Forward, we’ll call it – a lot of investor cash went into start-ups in Silicon Valley, and a lot of them put it into Silicon Valley Bank. The bank sensibly put a lot of the cash into long term bonds – safe as houses, unless… interest rates start shooting up. Then you have to sell some of the bonds at a discount and someone notices and says “get your money out of the bank”, which of course causes a run on the bank and – crunch – it falls over.
The FT reports:
As an institution that is estimated to work with half of local tech start-ups, its collapse presents an obvious threat. It led the head of Y Combinator, San Francisco’s pre-eminent accelerator for early stage tech companies, to warn on Friday that Silicon Valley’s start-ups could be facing an “extinction-level event”.
Which is exactly the sort of calm, considered language that responsible business people need to use to avoid panic and cause a bank run.
Stay tuned until Monday to see whether Silicon Valley adds to its recent humanity-benefitting deficit by triggering a full-on global financial crisis…
Tome, to you
Tome is an AI presentation generator or "story teller". Uncanny.
I’ve only used this once out of curiosity but oh my gosh, it was good. This is a presentation I made from a prompt about the arctic explorer Amundsen and his focus on learning. It generated all the text and some half decent images to go with it…
Exploring Amundsen’s Learning Legacy
It takes an AI to spot an AI
Human writer or AI? Scholars build a detection tool, Stanford Human-Centered Artificial Intelligence
DetectGPT can determine with up to 95% accuracy whether a large language model wrote that essay or social media post.
Very much like GPTZero – which lets you have a go for free and worth using if you think someone is giving you unfiltered machine generated copy instead of spending valuable hours writing low-grade prose under their own steam.
Bing is better (for now)
Many colleagues have mentioned to me they’re using ChatGPT for all sorts of tasks Some of them have sounded more suited to Bing – things like research or to do with topics which are very current (ChatGPT’s reference data cuts off a year or two back.) Microsoft’s Bing search engine has been disregarded, mainly because people tried it years ago and found Google better.
Now though, Bing is amazingly useful. Current data. Links to the sources and way fast than wading through pages of Google results.
Here’s an example of a brilliant response from Bing. I was thinking about “mic drop” sentences in novels, like “Reader, I married him” in Jane Eyre. How do they work technically?
The answer is comprehensive. But also I was able to continue getting deeper results by asking more questions of Bing. Using the current verison of Google, some of the same results came up that Bing used to form its answer. But so much easier and faster as a whole answer.
That usefulness may soon degrade, as Private Eye reports that Microsoft is briefing companies on how they might pay to be included more prominently in the search results – the same force that has created search results on Google that are about as useful as pile of discount leaflets falling out of a free newspaper. Also, new developments in AI are going to keep coming at a fast rate this year. A new generation of the tech behind ChatGPT is rumoured to be coming out this week; and we haven’t even seen Google’s AI-powered services yet.
Like mega lists of trends, these aggregations of apps and information about artificial intelligence cater to an understandable FOMO a kind of data kleptomania.
I don’t know everything is going on, but I do have this very long list of everything that is going on next.
But if you are ready to commit some hours of reading and research, these can be a good way to start, although it helps to set some filters and intentions as you set off.
Here’s one via Ian Crocombe that looks good: The ChatGPT list of lists: A collection of 3000+ prompts, examples, use-cases, tools, APIs, extensions, fails and other resources
There’s also a big database available if you sign up and then recommend to the high volume, enthusiastic and useful Ben’s Bites newsletter, which is produced by Ben Tossell of Zapier.
See your Wifi deadzones!
I both love XR (virtual reality and augmented reality) tech and am deeply skeptical about its current use cases. The tech has been there for a while and there are have been three surges in “it’s actually happening” hype in the last decade or so. This app on the Oculus (Facebook’s googles) looks actually useful – it let’s you “see” where the dead-spots are for wifi in your home.
Meanwhile, three miles below the surface…
Most Saturdays, I take part in my local parkrun. Here’s a shot of some of the all-women team of 20 pacers as part of the celebration of International Women’s Day this week.
All parkruns are 5km, which takes me a bit over half an hour to get round the course at the moment. The very fastest runners who take part come in between 15 and 20 minutes.
Imagine that 5,000 metre run laid out in a straight line. Now imagine it going down into the Pacific Ocean, past the sea bed and into a chasm, deeper and deeper until you reach the bottom, or abyssal plane.
That’s where Adrian Glover and his colleagues are researching the weird and wonderful flora and fauna.
Follow their work and see more images likes these via his Twitter feed.
How I work, by George Tannenbaum
I. Love. This.
A brilliant story about the creative process as well as an example of pitch perfect content marketing in itself. 16 slides with very little on them from copywriter George Tannanbaum.
Together alone with 50,000 friends
This is either a tragic story of trying to replace something that is still there for the having – the office – or a fascinating story of accidental innovation by crowds.
TikTok and Zoom body-doubling – live-streaming someone working at their desk – makes people feel less lonely and gives an ambient prompt for them to focus on their work, according to a Fortune article. We run in groups, we live in groups, and we often need work in groups even when we’re doing tasks that seem solitary.
“If you observe a whole screen of people focusing and working, it’s much easier for your own nervous system to calm down and to almost subconsciously mirror those positive behaviors,” Navarro says.
Despite Elon’s best efforts, there’s life in the old Twitter platform yet. Here are three superb examples of Twitter serendipity serving up useful information.
1. Sky Sports news presenter nails the weirdness and hypocrisy of suspending of top UK football presenter, Gary Lineker.
2. QR codes were laughed, became useful and then ubiquitous. Then the Prime Minister’s communications team reminds what stupid looks like:
Well done sleuths I confess I saw it and could not be arsed to jump through hoops
3. Now this is a perfect example of a well written Twitter thread – Morning Brew explains the timeline of how Silicon Valley Bank blew it in just 30 hours –essentially a series of poor decisions made too late.
The Creative Act, by Rick Rubin.
A nourishing book with its head in the spiritual but it’s feet planted suprisingly firmly in reality. Maybe because I’ve read a lot of neuroscience-related material in the past few years, I find lots of connections in his reflections on being an artist with scientific terms:
If you have an idea you’re excited about and you don’t bring it to life, it’s not uncommon for the idea to find its voice through another maker. This isn’t because the other artist stole your idea, but because the idea’s time has come.
This is known as “simultaneous invention”. Darwin wrote Origin of Species while Alfred Russel Wallace was writing essentially the same thesis. Films about remarkably similar subjects arrive in little waves. There’s a mystical sensibility to the notion that some ideas are just “ready to be born’. This observed phenomenon may have its explanation in systems theory, complexity and emergence. We’re used to thinking ourselves as individuals, when in fact we are part of massive systems of minds and memes that produce new things from works of art to practical innovations. There may be muses whispering in our ears and Rubin thinks of those voices as “the universe”, while I think of them as cultural systems, interconnected networks of signals and meaning. Maybe that’s the same thing.
That’s all for this week, folks.
Thanks for reading, liking and subscribing!
While I do appreciate having a “sensory overload”, a “information overload” won't be needed …
As someone who is [hopefully] heading up the slope of enlightenment, I seem to fall back into the valley of despair on a regular basis