This is a good take on the viability of generative AI businesses.

Except this bit:

Handily, consumers are really picky about what goes in their mouth. The unofficial motto of your main competitor is “Is Pepsi ok?”. This is despite the fact that they are identical in both taste and colour. And a significant minority of people actually say no!

These drinks all absolutely taste significantly different. That’s why it works: people are picky about taste and the flavours being distinct.

In plain terms, everybody is being [messed] with constantly in tiny little ways by most apps and services, and I believe that billions of people being [messed] with at once in all of these ways has profound psychological and social consequences that we’re not meaningfully discussing.

As usual, long and with many swears, but I agree with much of this “Never Forgive Them” post (apart from, ironically, that things are unforgivable).

Growth is such a bad driver, and VC is one of the biggest rotten cores.

Also this:

to push users into doing things that either keep them on the app longer

I had this argument so many times at N-able, where there was a pervasive idea that more time using the product was good (ostensibly for the user, but really for revenue), and I continually argued that less time was, in fact, better.

Handy package to convert various types of file to markdown, via DF.

Most interesting is that this is mostly a wrapper around other tools, and which ones Microsoft has picked.

Neat gift idea:

a list of some of his favourite[…] five books, five music videos, five TV shows and five movies […]. Whenever I watched one of the recommendations I got to talk about it with him. I’ve also had success with giving a friend a voucher promising to read a book of their choice in the next year, because one of the worst feelings is reading or watching something amazing and then not having anyone to talk to about it.

From the Spinoff gifts without buying guide.

Nice set of ‘basics’ for someone working in software development.

Interestingly, I agree with the conclusion of this hire two junior devs or one senior dev question, but for totally different reasons.

However, if you hire strong early career engineers, some percentage of them will turn out to have the potential to become those 10Xers. If you can retain them over time, they will grow to lead the game-changing initiatives of your company, being both technically capable and fully embedded in your company’s culture. Those years growing up with you give them an internal network and high baseline of trust that is so critical for success when you’re talking about big technology projects.

💯

I do agree that titles have degraded in meaning over time, but was there ever a time when they were really consistent across the industry?

I had conversations like this:

Either you run in AWS/Azure or you build a data center, you string ethernet cables, you buy a backup generator, you hire a couple of engineers to keep these running when hard drives fail. What? No. This is not the alternative.

so many times. People seem to miss the middle ground.

I’m not sure I like the “stack native” term, but absolutely I would continue to build like this.

Very nicely visualised story about how songs about love have changed over time - I’m not sure I agree all these categories are “love songs”, so it seems to me the proportion has declined, but not by a huge amount.

I spent ages reading the Spinoff ranking of Auckland malls waiting to get to my local, only to find it ranked first place!

Thought-provoking piece on password hashing, with some history, and an argument that hashing prevented people realising for a long time how bad people’s passwords were. I’m not sure I totally buy that, but maybe, for some definition of “bad”.

Also an interesting thought that passwords should be asynchronously encrypted instead of hashed, allowing research work using the private key. It’s right that you could never get approval for this now, but maybe it would be ok if there was the theoretical room where the keys were stored.

I don’t find this Paul Graham piece on writes and write-nots compelling. For one thing, I think it’s more “write well” than “write”, but that’s not clear. Either way, I’m not really convinced AI is actually going to do this well enough that some aspect doesn’t remain common. Even if “writing well” ends up meaning more of editing and prompting, it’s still a skill people will need.

Fun post on async functionality in programming languages - which is a spoiler, sorry, but it was pretty obvious what “red” and “blue” were very early, and that doesn’t ruin it.

Hard-coding things you can’t mention will be interesting to see scale. (Via Simon Willison)

Worth reading and thinking about: motivation and work - the waste aspect it brings up is so infuriating and it really does need people to just care.

Interesting (and deep) thoughts on async/await both in general and in specific languages.

Good comments on working in tech outside the Bay Area. The thoughts on the future of SaaS are particularly interesting (or least obvious to me, compared to the rest).

As solar power trends towards free or negative cost in places, this kind of practice where you adjust systems automatically with available sun/power will get more and more common, I expect.

This extract data by making it a video and using AI to analyse the video technique is clever and impressive, but also horrific and inefficient and I really hope tech turns back from this future.

“Many rank-and-file engineers have stories about submitting their resignation, or threatening to quit, and their managers plying them with stock or cash or promotions to stay.”

I’ve had this happen to me three times. One worked out really well, and we were both happy for years after. One worked out ok, but didn’t last as long. One left both sides less happy (but with me earning much more and the company keeping me for a while longer).

A lot of good advice (for both sides) in this long post on how long to stay, and how hard to work to retain.

Interesting explanation of why Apple autogenerated passwords look the way they do.

“This proclivity for romanticizing the past and growling about the present appears to be (a) perennial and (b) cross-cultural. It’s also (c) dangerous because it prepares the ground for reactionary practices. That’s true most obviously in politics, where demagogues weaponize nostalgia” Alfie Kohn

Interesting details on how Copilot prompts for suggestions in VS Code - potentially outdated now, of course, butt insightful into what’s happening under the hood.

I really enjoyed this talk on local-first software and ‘barefoot developers’. I agree with a lot of it, although I’m not sure whether generative AI will really have the impact people think it will (plenty of tools were meant to in the past). The call to action, having tooling that supports generative AI builders to guide where it goes, makes sense, though.

(Generative AI has so many other ethnical, environmental, economic, etc concerns as well, so do reasonable uses even survive the bubble?)