Air Bee’n Bee
One of my boys and his partner gave me a ‘bug hotel’ for my birthday and we decided to pin it to the garden fence and await developments. So far, five Mason Bees have booked accommodation.
Quote of the Day
“Humour can be dissected, as a frog can, but the thing dies in the process and the innards are discouraging to any but the purely scientific mind.”
- E.B. White
Musical alternative to the morning’s radio news
Eva Cassidy | Autumn Leaves
I know I know, it’s not Autumn yet. But this was new to me and I liked it. Hope you do too!
Long Read of the Day
Gordon Brown on his experience of the Murdoch press
Compelling piece in the Guardian by the UK’s former Prime Minister, in which he details what these rags had been up to — including deleting millions of potentially embarrassing (and perhaps legally problematic) emails.
The kicker is that the guy who presided over much of this chicanery has been appointed Editor of the sainted Washington Post. Truly, you couldn’t make it up.
Books, etc.
Human Voices
I’m currently reading Penelope Fitzgerald’s biography of her uncles, The Knox brothers, and am deeply impressed by it, particularly by her ability to provide contextual detail about four men who had complicated and interesting lives (one was an early cryptographer and code-breaker; two were priests; and one was Editor of Punch).
I hadn’t read anything by her before, and indeed only got the ‘brothers’ book because of reading Henry Oliver’s Second Act and discovering that she was fifty-eight before she published a book and yet is now recognised as one of the best English writers of her generation. So I started wondering which of her novels I should start with. Sarah Harkness’s recent essay solved the problem: Human Voices it shall be.
Which is why her essay is worth reading.
My commonplace booklet
I’m a sceptic about AGI and deeply suspicious of the giant tech corporations which aspire to control the technology. But from the outset I’ve been a pragmatist about Large Language Models (LLMs) like GPT-4, Llama, Claude, Gemini et al. I see these as (a) what Alison Gopnik calls (cultural technologies — like libraries, language, books, etc.) and (b) as potentially useful tools — like spreadsheets, and in that context use them a bit as unpaid but assiduous interns. And I find them useful.
Which explains why I liked this post by Nicholas Carlini. He’s a security researcher and a sceptic about most things, but in the essay he outlines how he’s been using LLMs in his work. His view is that,
current large language models have provided the single largest improvement to my productivity since the internet was created. Honestly, today, if you gave me the choice of solving a randomly selected programming task from my work either with access to the internet or access to a state of the art language model, I’d probably pick the language model more than half the time.
Which is interesting, is it not?
This Blog is also available as an email three days a week. If you think that might suit you better, why not subscribe? One email on Mondays, Wednesdays and Fridays delivered to your inbox at 6am UK time. It’s free, and you can always unsubscribe if you conclude your inbox is full enough already!