ttiurani

Solid point.

A bit tangential, but I've recently been baffled as to why is this belief held by so many:

"since LLMs are trained on online data, they are re-trained with their own AI slop, further damaging them. Soon bad code will get even worse, basically."

together with:

"I’m sure it will get better with time, but it’s not there yet."

What improvements people think will overcome the poisoning of the well? I'm not an expert at all, but to me it feels like we'd need a new breakthrough to get good output from garbage input.

I realized I have aphantasia, not SDAM, just a year ago or so. Mine is mild because I have visual dreams, but just can't get anything going when I'm awake.

I'm now playing around with visualizations just when I'm falling asleep, when I now notice I actually can do it. To me, seeing pictures in my head just feels very odd and kind of pointless.

"You don’t believe in IPR. Then shut the fuck up about IPR."

As a software engineer, I need to call out this rhetoric. We are not all morally bankrupt. Many of the people with the most ethical backbone I know are also software engineers.

I don't use "AI" for the same reason I don't use amphetamine. I'm sure I could get stuff more stuff done faster if I used it, but I don't, because the manufacturing and the distribution of it is completely unethical.

The blatant stealing of intellectual property is only the tip of the shitberg. As bad are the giant ecological footprint[1], the colonialism of the grossly underpaid RLHF, and the conceding of even more of our digital lives to trillion dollar corporations.

We can and must ethically evaluate our work, and that includes our tools. Nobody's perfect, but doing one immoral thing does not mean you get a free pass to skip morality altogether.

1. https://www.greenpeace.de/publikationen/environmental-impact...

On the same boat with 25 years of programming behind me, and similar thoughts.

My 3 counter points to OP.

1. Sleep as much as you can. Looking back I have produced the best results when I've slept close to 9 hours. Too much sleep is almost never a problem because I can't produce high quality code for more than about 6 hours per day. When I try, I just tread water, because I have to fix what I did the day before. (If you have more than 10% monotoneus coding you can do almost asleep, you haven't yet fixed the code to eliminate that work, or are in the wrong place.)

2. Try to get to a place where your "career" doesn't matter. This is a privilege that may be impossible to get, but if you can, try to be "just" a team member many times during the years, no matter how high up the ladder you've ended up earlier. IMO that's the best way to keep learning and improving in the craft.

3. Most importantly, read, understand and internalize how you impact to the world as a technologist. Get a deep understanding about society, ecology and ethics related to technology. Know where the money flows, keep crystal clear who gets wealthy and what gets destroyed to generate that wealth. Aim to get to a place where your code is net positive for the world based on that knowledge.

Indeed, this is a vexing problem.

Role models influence so much of peoples' behavior, that I fear they are needed to change what is considered success. But the people suited to be role models rarely want to be.

I don't know the answer, but hope some of these kinds of people would allow themselves to be (reluctantly) also celebrated.

> All around us are these lives — heads down and arms open — that ignore the siren call of flashy American individualism, of bright lights and center stage. I’m fine right here is the response from the edge of the room, and that contentment is downright subversive. How could you want only that? the world demands. There’s more to have, always more.

Beautiful writing, and I feel this part especially elevates it.

Going forward, we collectively need to recognize and celebrate these people who know when they have enough for a good life. Who can stop craving more fame, wealth and possession, and just appreciate what they have. Because only by doing so, can we leave enough resources for others, near or far, now and in the future, to have the same.

For the past few weeks I've been using

https://twiiit.com/

to find independent nitter servers that can show tweets of the few academics I still want to follow.

Now Twitter might soon break also these small third party servers but for the moment, they still work.

I'm running the last(?) Intel MBP from 2019 for rust and web development and haven't found any reason to upgrade.

Interesting observation: my remote junior colleague has a M2 Pro MBP with more memory than me. We're running the same OSX and node versions. Spinning up a development environment in a project using vite and a Java DB process running inside docker is about 3x faster on my old Intel than on their M2 Pro. The problem isn't anything trivial, and I'm struggling to help them figure it out over Zoom.

Not saying it isn't fixable, but I suspect there are still ARM potholes you trip on.

Just like individual flow, group flow is not an automatic and you can do great things without it. Its role also is not to keep the group "together".

> By definition, a flow state is non-collaborative; [...] you’re certainly not going to be in that mindset while you’re in a pair programming session.

This is not true. Group flow is a well documented phenomenom[0]. While it's easiest to recognize for example in an orchestra, group flow can happen with pair programming as well.

[0] https://www.aalto.fi/en/oasis-of-radical-wellbeing/group-flo...

I reached a similar conclusion with my own blog[0] and run a small rust process that parses access logs.

However, I could only find a need to know which posts are popular and which give 404 errors, nothing else. I'd argue that for big chunk of personal blogs, collecting general geographic and/or traffic sources isn't actually useful nor justified.

[0] https://tiuraniemi.org/blog/site-setup#_statistics

These are "words that deserve wider use". If this site is successful then some of these words become popular, and eventually people don't need to look them up anymore.

I write about environmentalism and technology.

https://tiuraniemi.org/blog

What's very important to note is that a big bulk of the recent reductions have come from picking low-hanging fruits, i.e. by replacing coal with natural gas.

This does not mean that going to zero will be possible with just more of the same.

The teacher of course brings the material to the student, typically by presenting it by making statements. If the emvironment is right and the teacher has good pedagogical skills, that should spur questions from the students.

This reminds me of a gem I picked up in my pedagogical studies: the aim of effective learning is for the student to ask from the teacher to explain something they don't understand. That both guarantees relevance and is highly likely to leave a lasting impression.

However, in most classrooms it's the teacher doing the asking. "Why are you asking me if you already know the answer?" is a completely valid response. A teacher should try to avoid that and instead ask e.g. "what are you having trouble with?"

With regard to ChatGPT passing tests, IMO the root problem is the overemphasis on tests with these kinds of redundant questions (and constant grading) that have always had limited correlation with learning.

For the record, I agree with your criticism. I did generalize my experience without enough justification.

It was a 30 second runaway comment and the latter part clearly doesn't hold up against scrutiny.

I agree and that's why I referred to "the edited sentences" and not English in general.

A more critical issue with just electrifying cars without also reducing the number of them, is that it's completely inadequate to meet CO2 targets. A new study in Nature Communications modelled this for London and found that:

"[T]he current system cannot reach stringent carbon budgets without adopting highly aggressive and disruptive policies. Electrification, including moving the phase out date forward, results in cumulative emissions 7 times greater than the Tyndall carbon budget for the “well below 2 °C and pursuing 1.5 °C” global temperature target. Rather, a combination of aggressive policies is necessary so that future emissions reach levels comparable to the carbon budget. Of these policies, the most important is reducing car travel activity. Policies that decrease car distance driven and car ownership by over 80% as compared to current levels are highly effective in edging close to the designated carbon budget."

https://www.nature.com/articles/s41467-023-37728-x

> I'm going to impose my preference on everyone else, even though the term has been part of standard contemporary English for a long time.

At least I, as a non-native speaker, find the edited sentences always easier to read. They simply make the text better.

As the entire point od Wikipedia is to make knowledge accessible with co-writing, I find it just wild that people would object to better language.

So who exactly is imposing their preference on the world: the one making the text easier to read, or the one objecting to the edits?

Exactly.

Electrification of single-occupancy vehicles – which, let's face it, is what the overwhelming majority of the EV business is really about – is not ecological. When you take into account environmental stress in general[0], and don't tunnel vision just on CO2[1], it should be immediately obvious to everyone that stopping the climate and ecological breakdown needs much more imagination than just replacing a car with another car.

[0] https://goodlife.leeds.ac.uk/national-trends/country-trends/...

[1] https://www.insnet.org/carbon-tunnel-vision-or-inter-connect...

Have you tried to migrate from raze to crate universe? It's close to feature parity and raze seems to be close to deprecation.

I successfully migrated recently and it's such a gigantic improvement.

A major addition to this report compared to previous ones was the emphasis on "demand side" mitigations. That means scaling down excess energy use in a way that doesn't compromise people's quality of life.

For example this paper suggests that in the UK a reduction of 52% by 2050 compared with 2020 levels is possible.

https://www.nature.com/articles/s41560-022-01057-y

Raph Levien, who's behind Druid and Xilem (https://raphlinus.github.io/rust/gui/2022/05/07/ui-architect...) UI frameworks, recently gave a talk about very similar issues. Slides here:

https://docs.google.com/presentation/d/1P8Su5mZSYkfZ1A9mPAaK...

> [T]echnology is about taking what’s out there in the world, and harnessing it for human purposes.

Technology is indeed about transforming existing nature into utility and producing waste thrown back into nature.

Knowing what we know about current terrifying climate and ecological breakdown, it's astonishing that Wolfram doesn't include it in any way in his visions. Even at the current pace nature regenerates far too slow to withstand this "progress", let alone in the visions in this article.

New technology is a choice to destroy nature for utility. Are we sure that utility is worth it?

A team of 10 devs I was a part of got lots of everyday value from JSON schemas (w/ code generation) being the single source of truth of the domain model for a Typescript frontend and Java backend.

I would absolutely choose JSON schemas again for any complex enough project that chooses typed languages for both backend and frontend. (And JSON as the protocol, naturally.)

> As soon as students raised a hue and cry over some issue — any issue — the administrators just rolled over, as though they were not actually in charge of the institutions they nominally oversaw.

If the students and their parents pay the administrators' salaries, they are in charge. These clashes between students and faculty tend to be much more maintainable in countries with free public education, because the students aren't customers.

> A few thousand

It's around 5000 per hour, well over a million new active users this month.

https://mastodon.social/@atomicpoet/109368148529393065

The title is a bit misleading. (Maybe even deliberately so, given this is FT.)

This story is about the recent surprise bestsellers by Kohei Saito in Japan. While Saito does argue for "degrowth communism", marxism is not the only way to frame degrowth ideas, and there are many in the movement who are not marxists.