_puk

Haven't they already started to regress?

I'm bullish on specific areas improving (I'm sure you could selectively train an LLM on the latest Angular version to replace the majority of front-end devs given enough time and money, it's a limited problem space and a strongly opinionated framework after all), but for the most part enshittification is already starting to happen with the general models.

Nowadays even ChatGPT doesn't bother to even refer to the original question posed after a few responses, so you're left summarising a conversation and starting a new context to get anywhere.

So, yeah, I think we're very much into finding the equilibrium now. Cost vs scale. Exponential improvements won't be in the general LLMs.

Happy to be wrong on this one..

Dubai chocolate [0] springs to mind as a great example. It's everywhere [1] now.

0: https://en.m.wikipedia.org/wiki/Dubai_chocolate

1: https://www.lidl.co.uk/p/della-sante-dubai-chocolate-cream/p...

Thank you so much for WSD.. I've used it extensively for at least a decade. We even self hosted it at one place after I pushed for it.

It was indeed pioneering in its day, though I find the default mermaid integration in the likes of Notion make it the (generic) goto for the team now.

Is that what the market demands, or is the market unable to differentiate?

From the site there's a huge assumption that potential customers are aware of what extrait de parfum is vs eau de parfum (or even eau de toilette!).

Might be worth a call out that these fragrances are in fact a standard above the norm.

"The highest quality fragrance money can buy" kind of thing.

Sounds good.

What languages do you support?

Learning Latvian through Anki flashcards, but it's not well supported by the main platforms, and there's not a huge amount of content out there for learning.

This alongside a couple of the usual suspects.

As a side note, on a Pixel 4a 5G (old phone , but functionally not ready for e-waste) the homepage bleeds all over. Some components into each other, others off screen. Might want to check that.

Something on my ideas list was some interactive game type thing for when cursor is busy "thinking".

This should fit the bill nicely!

Love the ebike analogy!

You're right though, you need to be able to steer, but you don't necessarily need to be able to map read.

Case in point, I recently stood up my first project in supabase, cursor happily created the tables, secure RLS rules etc in a fraction of the time it would take me.

To stop it getting spaghetti I had to add a rule "I'm developing a first version - add everything to an SQL file that tears down and recreates everything cleanly".

This prevented hundreds of migration files being created, allowed me to retain and context, and every now and then ask "have you just made my database insecure", which 50:50 resulted in me learning something, or a "whoopsie, let me sort that".

If I wasn't aware of this then it's highly likely my project would be full of holes.

Maybe it still is, but ignorance is bliss (and 3 different LLMs can't be wrong can they?!)

I've shared this with friends that are teaching, and they're definitely going to try it.

It was but 4 weeks ago that I had a very animated conversation with a teacher friend who wanted me to help them simply "just make AI do it" when they were doing lesson plans at short notice.

It's not just naive either, a few of them have gotten Google code to a point (with GPT direction) whereby they are able to make a presentation, but it's not quite fit for purpose, and as we know the last 10% is 90% of the work.. so it never quite sees the light of day.

In Wales (UK), they've leant fully into "personalised learning plans". Not to mention they've changed the curriculum completely in the past few years.

This essentially means that every lesson taught has to have a learning objective that meets the needs of each student, so there's a ton of overhead in tweaking, even if it is something that has been taught before.

Great to see this.

Ironically, given Google's stronghold over the past decade, I strongly feel that one of the big winners in the AI space is going to be the backend search engine.

Modern web search has become so polluted, with many tricks to get to the front page of Google that a lot (most?) of the good content is lost.

Now that many of the big models are capable of calling out to the web, this bloat is now appearing in AI search. A proper data first engine, without ads, less focus on presentation, and more on structured data is what is needed.

Sites are ranked higher when they have no ads. Fully open source.

That's a good starting point..

This is great to see, thanks.

There's so much time spent (dare I say wasted) on building and tweaking materials by teachers that it feels like an area AI can make a huge impact, directly improving the quality of life for a large number.

Not sure what it's like in Australia, but in the UK many teachers are expected to create materials in their own time (or just don't have enough time in the day to do so).

Oftentimes it's something that's been taught previously, but the teacher has a learning objective in mind, and so needs to tweak the existing presentation.

Love seeing some focus in this area.

I agree about the hidden work, but you describe it in a reactionary manner.

A huge amount of management effort goes into being proactive so you don't even need to get to the point in the conversation where they voice wanting to leave. Or when they do, you've seen it coming, have plans in place and they move on with your blessing.

* Creating a culture where employees (and managers) feel valued and have autonomy.

* Ensuring employees are clear about where they are headed, giving them something to work towards.

* Proactively identifying and unblocking issues before they become crises, often across functions.

* Identifying areas of frustrations for teams and individuals.

That's just a small number of the things you'll be doing, and whilst scope varies massively by company size (though until you get to the large corp environment most of this is still covered up to the VP / CTO level), you'll be juggling it all alongside technical input, cross functional leadership responsibilities (VP may well be in the SLT if the company isn't huge), horizon scanning, constant process refinement.. The pillars of people, process and technology all need to be constantly considered..

Having done this repeatedly from IC all the way through Head, Director and VP level (and up to CTO), I can say that being a proactive problem solver is always required of a good manager / Director / VP.

Edit: Having read the actual article ;)

IC paths map this quite well too..

Tech Lead maps to manager, Staff maps to Director, Principal to VP

Lol, not sure where I got the 2k from. Brain fart, but I'll let it stand :D

Back up your $20 a month subscription with a $2000 Mac Studio for those days your internet is down.

Peak HN.

I thought FB marketplace had peaked already, especially for furniture.

18 months ago you could find stuff on there fairly easily, now it's full of "must go" or "Free" items that are "message me for pricelist" when you click through. It ends up being a real slog to find anything.

Having spent the past week deep in cursor, it's amazing for building out a basic web app.. but getting it to a point of productivity takes a while..

Command line install for latest svelte.. nope npx install is now deprecated, have to do it another way.. ok, let's go old school and read the docs.

Great, it's up and running, but nope, Svelte has just hit V5 and the LLM isn't aware of the changes.. ok, do I drop back to 4 on a new code write, or spend the time to get a decent .cursorrules in place to make sure it's using the right version. Ok, that's in, but look tailwind is too new too.. ok, fine let's get that into .cursorrules..

Oh look, daisy has a v5 coming in 15 days, luckily I got in there just in time..

I thought svelte, tailwind and daisy were the one true way nowadays!

I now have a rule in my cursorrules that asks for any errors that I spot in the code (related to wrong versions) results in both the fix and a rule for the cursorrules so it doesn't happen again. That works well.

Whilst the pendulum seems well on its way to be swinging from microservices back to monoliths, I'm thinking we'll end up in a place that limits the volume and complexity of the code in a single service so that it's just large enough to encompass a point of single responsibility.

Then we can easily drop in and out of using LLMs in the code space.

Service Oriented Architecture lends itself well to the limited context of these models.

Maybe we can revive literate programming and simply build everything from a single markdown document..

Ah, we're holding it wrong. Got it.

This completely..

You ask a rote question and you'll get a rote answer while the interviewee is busy looking at a fixed point on the screen.

You then ask a pointed question about something they know or care about, and suddenly their face lights up, they're animated, and they are looking around.

It's a huge tell.

Does the air Canada website work nowadays? Last time I tried (admittedly 2 years ago) I found myself literally screaming "just take my goddamn money".. the most painful website I've used for years.

And then they lost our luggage (tents and all), but that's another story.

Have to agree. Die hard Airbnb user for years, then recently using booking.com more and more as it's clear and obvious what you are paying for (in most instances), and no insane "we're not a hotel, so change the sheets, wash them and put the rubbish out in these specific bins" list of exit rules.

Accidental discovery, but have to say they've helped far more than hindered (especially with last minute bookings in distant lands).

Works on a Pixel 4A 5G..

Pretty sure that's not doing any fancy on-device models!

That said, there was a popup today saying that assistant is now using Gemini, so I just enabled it to try. Could well have changed in the last week.

Of course it does.

Think about how, even in software, knowing that a physical shipping of a product (CD / Blu-ray) can be updated from "day 1" has led to poor quality releases with last minute patches.

The cost of having to physically recall / resend CDs back in the day meant that what went out had to work. The cost of sloppy software has now been externalised.

"efficiency and productivity is clearly more important for development of all industries than what you value"

And yet countries that value more than efficiency and productivity are repeatedly ranked as the happiest in the world [0].

"Just push for progress in natural way, and for politicians to adjust to an already happend changes"

How much damage has unfettered tech done to society at large? Maybe we'll never know the full answer, but we can definitely agree it hasn't all been positive (advertising, social media etc..). We're now retrospectively regulating, but a lot of it feels too little, too late, for some elements of society (privacy being a notable example).

0: https://worldhappiness.report/ed/2020/the-nordic-exceptional...

I think the GP's point (badly made) may be that the Lord of the rings example is addressed explicitly in the article.

"titles containing a colon are split and either side counts as a title drop. So for The Lord of the Rings: The Fellowship of the Ring either "Lord of the Rings" or "Fellowship of the Ring" would count as title drops"

Have you seen vanilla js lately?

It's moved on massively as a language, picked up many, many new paradigms.

Going back to code more than a few years old and your newer devs are going to be asking "what's all this var nonsense"?!

"Lindsey cautioned that this is a preliminary finding since they only studied one type of disc and plan to conduct further experiments with other disc types. "Considering the lack of data relating to the various details of a disc golf throw to resulting performance, we hope that this study serves as a catalyst for similar research... at other institutions," he said."

But in a system that exists without tests (this is the real world after all), the current functionality is already enshrined in the app.

Adding tests that capture the current state of things, so that when that bug is uncovered tests can easily be updated to the correct functionality to prove the bug prior to fixing it is a much better place to be than the status quo.

The horse may have bolted from the barn, but we can at least close the farm gate in the hopes of recapturing it eventually.

Drag to resize