I bought a new Mac. There was a bonus from work and it was exactly the amount that a new M2 MacBook Air cost, which I took as a sign from the fates that it was time to replace my 16in Intel-based MacBook Pro. I can also sell some machines which I have lying around and no longer really use to effectively cover the entire cost.
Of course I still have (and use) the ThinkPad running Linux but I can’t get out of the old technology journalist habit of having one machine for each of the main operating systems, because you never know when someone might commission me to write something. No one actually commissions me to write anything these days, mainly because I don’t actually have the time to write other than stuff for work, the odd blog post, and my creative writing. One day I will get everything down to one computer, but that day isn’t today.
Early impressions though are positive and it makes me realise quite how compromised the machines which Apple made during the 2010s were. Prior to 2015, I used a MacBook Air – first an 11in version, then a 13in one – and then in 2015 got the new 12in ultra-thin and light MacBook. That computer was Apple’s first to use the much-loathed butterfly keyboard, which was forgivable on a laptop which was designed to be incredibly thin. But using it on the rest of the range was one of Apple’s worst mistakes in its history, because it made their best-selling computers horrible to type on.
The 12in MacBook got replaced by a 13in i5-based MacBook Air (horrible keyboard, underpowered) and then a 16in Intel MacBook Pro (expensive, improved but still crap keyboard, underpowered because Intel was at a low point).
The M2 Air replaces that MacBook Pro and it’s like night and day. Literally, because I bought the “Midnight” version which is a delicious shade of almost-black blue. It’s also far snappier than the MacBook Pro, completely silent and – IMPORTANT – has a keyboard which you can type on. I can’t say how much of an improvement this keyboard is, and when you spend much of your life typing that really does matter a lot.
This MacBook Air is the first design that I’ve loved since the mid-00’s MacBooks, with their chunky polycarbonate (who didn’t love the black MacBook?). It’s almost as if Apple has remembered to make the Mac a combination of loveable and functional, after a fallow decade when it really lost its way.
Meanwhile we went to a local village fate yesterday where Kim was the judge for the cake category:
The fruit cakes were, apparently, all of high standard. While Kim was judging I spent some time working on some fiction I’ve been playing with for a while, re-plotting and outlining a story which hasn’t been quite hanging together. I’ve been using Aeon Timeline to do the outline, because it features some nice capabilities around navigating the complexities of multiple timelines while integrating with Scrivener. It’s a complex piece of software and I feel like I’m only just getting my head around it, despite using it for over a year.
This week I have been writing…
I wrote a short piece about the limitations of current AI, which takes me back about 25 years. Before I joined MacUser and became a journalist, I did a PhD in philosophy, looking at the implications of Kant’s philosophy of mind for artificial intelligence. At that time, cognitive science – a blend of computer science, philosophy and psychology – was the hot thing in AI, but in the past 10 years or so there seems to have been a return to the AI of the earlier years, which attempts to subdivide “intelligence” into a set of discreet functions capable of being developed in parallel.
My old thesis basically said the opposite: consciousness is a necessary part of what we mean when we talk about intelligence as it’s instantiated in humans, and consciousness is unitary (there are many functions in the brain, but only one “I”, no matter if you’re a human, a monkey, or a lizard). No amount of bolting a vision system on to an abstract reasoning processor on to a large language model will get you to unitary consciousness.
Was I right? I think the fact that despite the best efforts of very smart people, we are no closer to creating animal-like consciousness probably means I was. Large language models are impressive, but they “know” nothing – saying they do is a category mistake, as Gilbert Ryle would have put it.
This week I have been watching and reading…
We binge-watched What we do in the shadows this week and I haven’t watched anything which made me laugh out loud so much for a while. We’ve been wandering round the house randomly shouting “BAT!” in a Matt Berry voice, then collapsing into laughter. With Ted Lasso and The Mandalorian on at the same time, and The Power also now out, there’s plenty to watch.
Two new books on the virtual book pile: Ken McLeod’s Beyond the reach of Earth and Katherine May’s Enchantment. I enjoyed McLeod’s book is the second in a series, the first of which I enjoyed quite a bit, and I greatly enjoyed May’s Wintering too (pretty much a lockdown book) so I’m looking forward to both. But first I need to finish Becky Chambers’ Record of a spaceborn few which I have been dithering over for a while.
