I could as easily have subtitled this post “Who’s afraid of the big bad wolf?” so as to join together our being in one of our favourite places in the country – the Roseland peninsula in Cornwall – and linking it with AI. One half of the couplet represents all that’s good and the other the spawn of satan.
We arrived here completely knackered after days of work on the allotment getting it ready for our temporary absence and a couple more days of readying the campervan for its inaugural long drive after the troubles (new engine, clutch, Uncle Tom Cobley and all. Up at five am only to discover that the roof had leaked in a place it had never leaked before – we quickly sorted it and set off in gathering sunshine for a long drive at a stately sixty in honour of the engine; “she being brand new and consequently a little stiff “. That last bit was a quote from a poem by ee cummings which a musty old critic once described as only fit for adolescents. I haven’t aged a day!
Musty old critics have had their field day over AI as well – marching en mass under the banner of the goose quill pen and denouncing AI vehemently, claiming to protect the masses from evolving technology. Good luck with that then, although Google seems to have messed up their latest software update on my watch which insists that I only slept for three hours last night. It was ten hours and I felt all the better for it.
So part of the purpose of this trip is to field test a new plant recording system and that’s where the AI comes in. Knowing a place like this really well has its advantages for rehearsing a new way of plant recording; not least that I’ve already put in the hard miles for most of the species. But anyone who gets interested in plants and wildflowers will know that there are some very common species that are extremely hard to tell apart. So part of the aim is to revisit the difficult ones and see whether I can distinguish the trees from the wood. In preparation I picked three brain teasers and posed a question on Google Gemini. “How do I distinguish A from B” . The first one sounds easier than it is in practice. How do I distinguish between a wild Carrot and a sea carrot? In a few seconds the AI answer dealt with two things – the hows and the whys of wild carrots v sea carrots. The how is demonstrated in the photo at the top. It’s a wild carrot growing on a clifftop near here and that’s because the flower head is in the process of forming a basket/ lobster pot shape. But then you might well ask – what’s an ordinary wild carrot doing on a sea cliff? and the answer is that wild carrots normally grow in and around arable fields and sea carrots grow near the sea. BUT, growing close together, they readily hybridize and so in Cornwall where the arable fields often come to the clifftop with only a footpath between and so there are a lot of what we might call cousins that look somewhere between the two. My mind is at ease because the hybrid plants can be assigned to the (very large) category of WOB’s the walk on by’s.
The other two questions were necessarily more general – what specific details should I record with photographs and all the other tools – hand lens, ruler etc in order to distinguish and identify the other big families grasses (Poaceae) and the huge family of Dandelion/ daisy lookalikes. In a few moments Gemini gave me a brilliant crib sheet to take out on my adventures. Traditionalists may take a couple of minutes here to rearrange their scowls in case the wind changes and they get stuck with lemon faces. Each and every record will be taken home to my study and interrogated with the books.
Until the plant apps bed in – which they soon certainly will, the old way is still the final arbiter; but I can foresee a day when taxonomy will be a matter of taking a clipping from a specimen with a hand-held DNA analyser. How you might challenge an identification after it’s been deified by its genes by a person in a white coat who’s never as much as sniffed a rose is not one I’ll live long enough to worry about.
The thing about AI is that we have to make the cultural move from knowing everything by rote towards asking better and better questions. The skills of the future will lie in knowing how to extract the maximum good science through a question that’s as sharp as a samurai sword.
We’re having a recovery day first and apart from lazing around and snoozing I’ve been looking at photos of the allotment plants as if they were our children, and reading John Wright’s excellent new book Grasslands. A further search with AI revealed that grasses are by far the most numerous species if you exclude the 57 varieties (OK thousands, then) of the apomicts which have dispensed with messy sex and just clone themselves; Dandelions and Blackberries are two of the other culprits here in the UK. The following two species in numerical order are the daisy (dandelion) family, then the Sedges and the roses. I can’t see myself getting bored anytime soon.
There’s not a great difference between learning to live a virtuous life, reaping the benefits of eudaimonia (flourishing) and learning to make better plant records. They both require constant practice – which is one of two reasons that we’re here. The other reason is the joy of walks near the sea. There is – I suppose – a third reason which is far less virtuous. The botanical societies tend to cling to traditional ways, and getting records through the minefields and the heavily guarded portals into the universe of scientific data can be a tortuous and Kafkaesque experience. I should know. I worked inside a conservative organisation (the C of E) for 30 years and I learned that these organizations will never abandon a regulation without a decade of bloodshed. I’m trying to develop a personal recording method that will shortcut the traffic jams by providing exactly what the gatekeepers are asking for. I’m not trying to change anything, I’m plotting a path through the maze.
