Splitting a Domain Across Multiple Bounded Contexts by Mathias Verraes @mathiasverraes.bsky.social verraes.net/2021/06/spli... Design and Reality : Essays on Software Design by Rebecca Wirfs-Brock, Mathias Verraes leanpub.com/design-and-r...

I do not fear the rise of superintelligence. I do, however, fear the rise of billionaires, organizations, and world powers who seek to use computing to maximize their power, influence, and control.

"Stay out of politics" is almost always an authoritarian statement. It attempts to shut down public debate (a cornerstone of democracy) in favour of the person making the demand.

are you displeased with today’s AI safety evaluation landscape and curious about what greater conceptual clarity, methodological soundness, and rigour in AI evaluation could look like? if so, consider coming to Dublin to pursue a PhD with me apply here: aial.ie/hiring/phd-a... pls repost

A particular thing works is useful information but not the same as proving that it's the ONLY thing that works

"European alternatives to digital services and products, such as cloud services and SaaS" european-alternatives.eu

Yes, let's fill our homes with robots that are controlled by a fascist megalomaniac trillionaire as if we have never seen a sci-fi movie 🙄

(Dude suddenly stops pushing his broom across the floor of the Death Star) Does this look like a big laser to anyone else

The Gell-Mann Amnesia Effect seems to apply even more for AI than media.

Our Call for Papers is closed! Thank you to all 691 applicants who submitted. This response was incredible. We've announced our first confirmed speakers, with more to come soon: buff.ly/jIPNthk

Someday corporations and technology companies will pretend they were always against this

We've scheduled a session! @rebeccawb.bsky.social & @mathiasverraes.bsky.social will discuss 'Critically Engaging with Models'. Learn to evaluate models and their limitations. Pre-reading their essay advised. Jan 28, 8 PM CET. buff.ly/LXXZsu2

1725: Why the fuck did I lose my left hand for the same wages I earned while being 10x more productive than last year? Who the fuck is benefiting here? 2025: Why am I losing my job to a machine that encourages kids to kill themselves and can't actually do my job? Who the fuck is benefiting here?

BREAKING: This is huge news, the EU's equivalent of the 🇺🇸Supreme Court's 2015 Obergefell v. Hodges ruling. 🇪🇺Court of Justice just ruled all 🇪🇺countries must recognise same-sex marriages granted in other member states. This effectively legalises gay marriage across 🇪🇺

This post was about the previous peace plan, but yeah, this shit again.

This is peak intellect. The CEO of OpenAI.

"One in every 25 participants at 2025 UN climate summit is a fossil fuel lobbyist" We are being scammed out of the air we breathe. www.theguardian.com/environment/...

All models are reductions, but if yours fits in a 2x2 quadrant, you might have reduced too much.

Nobody complains when you make assumptions and they turn out right

It's exactly how it started for me and I still can't believe I did it twice. Wonderful conference for speakers, even with less experience, very supportive team and nice people attend, so now you have no excuses :P

If you're new my work on models: verraes.net/2022/09/crit...

It's not that I don't agree, it's more like it's treated somewhat like a mystic text (we all nod solemnly at its profundity) and that it seems like we have little else to say about models. (The essay and talk you did with Rebecca is a rare exception.)

It is misused or misunderstood (I think) to create a narrative of absolute relativity of models.

There is also perhaps a certain facileness in how we talk about models which rubs me the wrong way, which has its roots in the multitude of models and their uses, and also that we seem to be naive about models in general.

It's no deeper than the observation that a linear model is not useful for non-linear systems. But an excessive faith in the relativity of models seem to get in the way of that understanding.

The wrong model is much wronger than the right model.

It's also that people seem to read weird things into it, like all models are equally wrong, or all models are equally useful, or all models are useful on the same parameters, or that models are uniformly wrong or useful irrespective of context. I could go on.

To evaluate usefulness, we can't do analysis, we have to put the model to use and iterate. That's the antidote to perfection and paralysis: exist the model to daylight and see if people use it in productive ways.

Well, yes, the quote implies that we don't need to seek perfection, and we don't need to represent all elements from reality accurately in the model. But the second half of the quote tells us to evaluate not perfection (elegance, symmetry, realism...) but usefulness.

Interesting, I always read it differently - to remember that there's something "as good enough for my situation/context" to avoid analysis paralysis and looking for perfection :D

All of the sudden feeling an urge to go to Øredev

The quote "All models are wrong but some are useful" should not be read as an excuse to stick with your model. It's a call to actively search for more useful models.

That ceasefire lasted about as long as it takes to reload the canons.

Related to the paper we discussed at the most recent Papers in Systems meetup, "Critically Engaging With Models" by @rebeccawb.bsky.social and @mathiasverraes.bsky.social wirfs-brock.com/rebecca/blog...

What if an advertisement was your girlfriend?

I think the thing that is different with chatbots is that they have been shown to have the capability of influencing people's beliefs and behavior. Obviously this isn't a linear deterministic thing, but still, not great.

Imagine being fooled into entering a hologrammatic prison filled with make-believe personalities that know everything about you and can gaslight you into doing anything from buying products to making up a spiritual delusion to killing yourself

I added three new tips to my article on "How to Read More Books". verraes.net/2012/12/2012... I've been growing that list since 2012!

Bully logic: "I will stop beating you, unless I feel you are a threat to me". That's not a peace plan, it's an excuse for the next genocide.

Vibe codin' plus Ballmer peak equals what?

Today is a day when arts degrees are worthless, but the product of those degrees is so valuable it would kill an entire industry if they were made to pay for it.

Sometimes the constraint is Mb/s, sometimes it's WTF/s

@thunberggreta.bsky.social has done more to prevent and punish genocide than all 193 States in the UN, especially the 153 signatories of the genocide convention. Let that sink in

The only surprising thing to me, is that so many people still believe that Apple is somehow different from all the other billionaire-owned capitalist tech monopolies. These people always side with power.

Google spokesperson Castaneda: "allegations concerning YouTube are "simply not true," in part because it's a streaming service & not a social network" audit results: "YouTube’s algorithm violates the platform’s very own policies, recommending misinformation, violent content, hate speech & scams"

To be honest, given time and access to the people who vibe coded it, I quite enjoy cleaning up codebases that are a mess. Maybe I'm weird. But usually, having to deal with people who don't understand the value of the cleanup are what make a job like that shit.

I was told today that "we don't need to experiment, we don't have a problem". 🧵

The “et al” is such an amazing community! For those who weren’t able to join us: the paper is a valuable read and the discussion was so vibrant and engaging, we were a minute over before I even looked at the clock :D Thank you @rebeccawb.bsky.social , @mathiasverraes.bsky.social and everyone there

This was fun, thanks for having us @ruthmalan.bsky.social et al 😀