“Devs gaining little (if anything) from AI coding assistants | CIO”

https://www.cio.com/article/3540579/devs-gaining-little-if-anything-from-ai-coding-assistants.html?amp=1

> Many developers say AI coding assistants make them more productive, but a recent study set forth to measure their output and found no significant gains. Use of GitHub Copilot also introduced 41% more bugs, according to the study from Uplevel, a company providing insights from coding and collaboration data.

Like I've been saying for a while, feeling productive is not the same as being productive

Since every new release by OpenAI is usually followed by nonsense commentary and bubble inflation: a note

Testing a large statistical model with a handful of manual tests ("it solved this riddle!") is like assuming the home remedy you took cured your cold. These are methods that only lead to superstition.

“Kids who use ChatGPT as a study assistant do worse on tests | Popular Science”

https://www.popsci.com/technology/kids-who-use-chatgpt-as-a-study-assistant-do-worse-on-tests/

Do I really need to keep posting these? Haven't we all figured out that LLMs are extremely poor learning aids?

(He says, knowing full well that chatbots have become the default learning aid for coders.)

This doesn’t come as a surprise to most, but it definitely is surprising to many managers and consultants, but generative models—what’s being marketed as ‘AI’—are contraindications of quality and value

That is, anything those models do well is likely bullshit, so doing it more and better will make the product worse

The tasks and features generative models excel at were almost always bad ideas to begin with, and anything labelled as “AI” will have more of it

“Study Finds Consumers Are Actively Turned Off by Products That Use AI”

https://futurism.com/the-byte/study-consumers-turned-off-products-ai

> When AI is mentioned, it tends to lower emotional trust, which in turn decreases purchase intentions

Like I've been saying, we don't need the term "slop". Consumers have decided that "AI" in its entirety is bullshit. And, honestly, they’re right.

Please stop using generated art for thumbnails and illustrations for videos and blog posts that aren’t generated themselves. It gives people the impression that the post is spammy bullshit. Starting to see people dismiss links out of hand just because of the preview thumbnail on social media

“EXCLUSIVE: Naomi Wu and the Silence That Speaks Volumes”

"Now that they know that I could be dead in a ditch tomorrow and no one would give a shit or say a word I’m 1000x less safe here." https://www.hackingbutlegal.com/p/naomi-wu-and-the-silence-that-speaks-volumes

How would you design your app if you are worried that an authoritarian state might seize all of your user's data and work?

How would you design your website if you think the countries where your hosting datacentres are located might descend into civil war? Or get invaded?

Most of us who are working on web apps and websites today are doing so in an unstable world. The UK and US are fast becoming more authoritarian so it might not be a hypothetical question for much longer.

@turion Yeah, which are probably good ideas in any case as it would foster a more sustainable dependency ecosystem even if things don't turn bad.

@turion I mean, if you're working on an app or website _today_ and you expect the next twenty years to be extremely unstable, would that change the way you design the app or website?

Idle thought of the day: if you assume that environmental crises, political instability, and authoritarian states are going to become more common over the next couple of decades, how would that change the way you design your websites and apps?

I just published "I don't care how you web dev; I just need more better web apps" https://baldurbjarnason.com/2022/more-better-web-apps/

Anyway, I'm trying to rethink my approach to social media discourse with an eye towards being more constructive. Which of course led me to write a super-long overly wordy blog post 😄

Also:

"Write a short blog post this weekend, Baldur, you're still recovering from COVID", I said to myself.

Almost three thousand words later:

I clearly never listen to my own advice. 😝

Fields are rarely homogenous blobs with only one opinion. Odds are if you disagree with a field's orthodoxy, there will be a heterodox movement within it that's based on the current state of the field

Have a little bit of curiosity about other people's ideas. It's fun!

/fin

For example, a smart person, adhering to an ad hoc "I'm smart therefore what I see is true" process, reasoning from first principles, would come to the conclusion that the earth is at the centre of the solar system

It'd be next to impossible to change their minds with data

The effort to reason from first principles creates an emotional investment in the (wrong) conclusion that makes the conviction extremely hard to undo afterwards.

It becomes an different issue if you're in a position of power and influence and you begin to affect public policy or opinion. Which is pretty much where we're at with the US's current plague of dunce-savant tech CEOs

I can't change their minds so I'm appealing to the rest of you. Don't be like that. Approach fields with a genuine curiosity and try to engage with their history

Try not to pay attention to the tech types who are too far gone down this road. We'll all be better off for it

Now smartypants tend not to apply this to astronomy, because the story of Galileo is a powerful warning tale about exactly the problems with thinking like this. But biology, art, and sociology remain popular targets despite this.

Now that's normally fine. If you want to make shit up in your head about neuroscience, fine arts, or human behaviour when you're at home, kicking back on the sofa, and want to indulge in feeling like a smartypants, have at it. Enjoy yourself.

What these guys miss (and the tech industry misses because it's largely run by these guys) is that collective work using a structured process will trump high-IQ every time. It's the basis of progress in most fields: the scientific method, arts and humanities, construction, etc

Reading up one unfamiliar fields is actually a lot of fun. You can study and follow the arguments made in a field over the years and even come to a reasoned disagreement based on the current state of the field

Reasoning from first principles about unfamiliar fields is a popular past-time among many smart people in tech. They love to try to "solve" long-standing fields based on nothing more than their wits and personal observation.

But most fields have points in their history where they were stuck in false paradigms or fallacies because those were the conclusions a smart person would draw based on their wits and observation.

It takes years of collective work for fields to break out of these fallacies

The thing about people who have a lot of whatever it is that we're measuring with IQ is that they quite often tend to be way _more_ wrong than those who have less of whatever it is that those points stand for.

High IQ is kind of like superglue for opinions

Not only doesn't IQ compensate for ignorance or poor methodology, it acerbates the issue.

Rationalisation is a powerful drug. Most people can convince themselves of almost everything but those with high IQ tend to be _extraordinary_ at it.