Technology

I watched the latest Apple WWDC keynote presentation today and found it to be as exciting as mud. It’s effectively fixing and optimising everything from last year, and providing the kind of features they were promoting last year and failed to deliver.

What I found particularly grating was the revamped Image Playground which allows you to take a person from your photo library and use generative AI to put them into something or have them do something. This is somewhat worrying. Regardless of privacy of the AI model – e.g. whether it runs locally on device or in the “private” compute space, the problem is that this could potentially be abused. That said, it’s not as if people can’t already do this with the likes of Gemini’s Nano Banana Pro or Banana 2. Apple just makes it easier.

Speaking of which, Siri with Apple Intelligence will let users reframe photos and allow generative AI to fill in stuff that doesn’t exist because of the framing. Talk about Orwellian nightmares! Everybody will soon be part of the ministry of misinformation! It’s all fun and games until it isn’t.

Another thing I’ve noticed at work – there appears to be increased presence from something called “Applebot” which, I assume, is linked to Apple Intelligence/Siri. This bot (which legitimately comes from Apple IP addresses) was hitting a site I was examining pretty hard (but not as hard as others) which is somewhat worrying. Can we all expect these AI bots to start consuming web site bandwidth and CPU/memory as it crawls, digs and extracts information from websites?

Also, with all this Apple Intelligence baked into the phone and macOS, I find it strange that they haven’t mentioned about any possible improvements to call features such as call screening or scam detection. Are there any improvements? Even just having call screening as a toggle switch in the Control Centre for iOS would be a big improvement for me.

Everybody sing: Old McDonald had a server farm, A.I., A.I., NO?

Google, while I appreciate the efforts you’re going to make A.I. genuinely useful in Google Workspace, I’d really like you to revisit a few basic options that have gone untouched for decades. I’m talking about font selection and font sizes. At the moment, this is all you get:

In this case, I’ve chosen the larger size rather than normal. Normal, to me, is far too small when composing and sending messages. And what’s worse, you can’t specifically set a size for reading emails that have been sent in plain text. What are small, normal, large and larger sizes? Who knows – they are not allocated a number size – unlike the majority of font size options in.. oh, let me see, pretty much every other bloody product. Including Google Docs.

Whereas this is what I see in Outlook and Outlook for Web, I get a much better range of options:

Aptos 14 in Outlook is a good size. Larger Sans Serif in Gmail is larger than Aptos 14, though it is close. But it’s not close enough and annoys the heck out of me. As somebody who started their email journey on quite literally text terminals where there no such things as fonts or font sizes (as such). Everything was consistent.

Gmail has unfortunately been one of the few services that hasn’t really kept up with the basics – instead choosing AI over all else, and appears to be slipping behind even good old Pine which was a very capable email client back in the day. I just don’t get as much joy as writing emails with Gmail as I do with Outlook or any other client.

Reading email when the sender uses Gmail’s default small font or plain text is also a terrible experience. Take this email from my server to let me know local backups have completed. The font is too small. Gmail offers me no official method of setting a default size for reading email.

I’ve tried to create a Chrome extension using Claude that will override Google’s default fonts and font sizes, but while it works in principal, it’s an uphill struggle with potential compatibility issues and future updates from Google.

There is also a third-party Chrome extension that I’ve used in the past called V7 Gmail Zoom – and it works wonderfully well. You can set default text sizes for both reading and composing. But because Gmail ends out such tiny text by default, and having now gone “large” to make thing easier – email I send myself looks too big.

Sigh. This shouldn’t be a thing in 2026. Come on, Google, fix this stuff.

Bit of a strange week for postal mail. As we head towards the local council elections in May, I’ve been bombarded with letters from candidates. I’m sure the independent candidate tried to call the other day, but I was in the middle of a Teams work call. He left a leaflet. Then there were handwritten letters addressed to either me personally or “the resident” from the Lib Dems and the Conservatives, and a leaflet from Reform, which was torn up and thrown away immediately.

What amused me was a handwritten letter from a religious group – possibly THAT group that usually calls around when you least expect it (no, not the Spanish Inquisition – they’d be 192 years too late). I only knew it was some religious thing from an enclosed pamphlet about Jesus and God and things of that nature. The letter itself was indecipherable. The handwriting was so illegible I suspect it might have been written in ancient Aramaic (as opposed to aromatic, which conjures the images of delicious duck in hoi sin sauce).

On the electronic mail front, I had to sigh when the latest news of Directive 8020 from Supermassive Games, for whom I worked up until September of last year, had ended up in my Google Workspace spam folder. Given that I had set up the infrastructure for the marketing department to ensure maximum delivery (ensuring DKIM, SPF, and DMARC were all correct, etc.) for this sort of thing, it was hugely frustrating to find that Gmail decided to throw it in the junk because other people had been reporting those kinds of emails as spam. You can do everything in your power, technically, to get mail delivered, but you’re still at the mercy of a mail provider’s anti-spam/anti-phishing/anti-malware heuristics as to whether it will get seen by anybody. I signed up for the mailing list. It’s double opt-in (e.g. you get an email to confirm that you want to subscribe), but Google knows best. Anyhow, I’ve had to mess up my contacts list by whitelisting a no-reply email address because of Gmail’s design….

Anyway, Directive 8020 is due for release on the 12th May. I’m still undecided whether I’m going for the PC version (RTX 5070 Ti with 12Gb VRAM with Ultra 9 275Hx CPU) or the PS5 Pro (which apparently the game is optimised for).

I’m so excited to start the new job that I’m up at 4:45 am writing this blog post. I’m certainly going to feel it at the end of the day (though, to be fair, I did go to bed early and have had a good 6 hours of continuous, uninterrupted sleep)…

Anyway, the point of this post is to say that Apple’s new entry-level laptop, the MacBook Neo, is rather spiffy. For £599 (or £699 for double the storage and Touch ID), you get a MacBook powered by last year’s iPhone 16 Pro (and Pro Max) mobile processor (which, although sadly limited to 8Gb, will be more than enough for less demanding tasks). The chassis is aluminium (and comes in a variety of different colours), and the keyboard is full-size and sturdy. The display, while not as good as the MacBook Air or Pro models, is certainly good enough for most people. It hasn’t been calibrated to the many colour spaces that the more expensive MacBook range, for example. It doesn’t have MagSafe charging, but instead comes with two USB-C ports – albeit with one running at 10Gb/s thanks to USB 3 and the other at a much slower 480Mbs thanks to USB 2. Only the USB 3 port can be used to connect to an external monitor, if required. Therefore, a dongle that can drive Ethernet, video display, and multiple USB-A or USB-C ports will be part of the Neo owner’s arsenal.

But despite these limitations, this is a full, macOS-powered laptop for less than £700. It joins the Mac Mini (starting price of £599) as being one of the most affordable Macs ever made – and given the shitshow that is the AI industry is allegedly buying up all the RAM and NAND (SSD) chips for data centre use, this is an absolute bargain. Especially for the education sector.

I had intended to give one of my nephews my former SMG M2 MacBook Air, but alas, I was forced to sell it to keep things ticking along until I got the new job. But, given the price point of the Neo, when my finances are in a better position, I’d be more than happy to buy a MacBook Neo for him, as I know he wants to get into the computing field. While the Neo has neural processing built into the A18 Pro chip, I wouldn’t want to try running any LLMs on it. With the Swift programming language (all the tools are included), he could certainly learn to build his own Mac and iOS apps, for example. Then there’s Terminal. He could learn Unix (well, the BSD equivalent), too, which is a handy skill in today’s IT market (along with Bash, Python, etc.). As a result, it wouldn’t take him long to get used to Linux. There is huge potential here.

The Neo has a lot of potential. There are a lot of naysayers that say that 8Gb is not going to be enough – especially in five years’ time and it’ll just lead to e-waste – but as somebody who ran a 13″ MacBook Pro with the base M1 chip when they first launched with just 8Gb RAM and 256Gb SSD, it worked well enough for this sysadmin – though 16Gb or 24Gb are much more preferable for those going beyond word processing and light video editing work. Also, remember that Apple generally supports its products for at least 5-6 years, including OS updates. It will have to support 8Gb RAM on these models for that length of time. I have no doubt they may switch out the A18 Pro chip for the A19 Pro chip, which has 12 GB of onboard RAM at some point, but that won’t happen for at least a year or two.

I think the MacBook Neo will take off and do spectacularly well, given the current economic climate, and I look forward to seeing the first batch of reviews as they come in this week.