Antoine Alberti

Dev, socio-poetical anarchitect, agile/lean evangelist under cover, grey areas researcher, self-certified itDependser and rubberDucker
I can name things and invalidate caches
coach & trainer @ arolla

blog
blog.antoinealberti.com
Antoine Alberti boosted:
2025-11-19

Complexity can always be added later, usually at low cost. Managing complexity is an unaccounted-for ongoing cost that these decisions almost never take into consideration - often because the people making the decision have no long term experience of managing the thing being added.
The cost-to-change of the thing being added is also almost always never accounted for in the decision-making, again because the people making the decision almost never have the experience of the actual work required to remove it from a running system.
Make sure your architecture decision-making checklist includes these two questions:
1) After we add this thing, how much will it cost to maintain on an ongoing basis?
2) What will it cost to remove?
If neither of these things can be accounted for, defer the decision until real-world data can be found.
Go ask in your communities of practice.

Antoine Albertiantoinealberti
2025-11-19

@interfluidity
Well at least I would like to see whether a national ad distribution service can operate ethically, or at least operate while respecting GDPR for real

Antoine Alberti boosted:
kulupu jynjyn@tech.lgbt
2025-11-19

it really concerns me that people treat AWS and cloudflare as infinitely available, such that the internet does break when they go down. google does a thing internally where they intentionally add brownouts so people don’t rely on services always being up, i’d like to see something like that model for the public internet.

and to be clear, I don’t just mean this for infrastructure, I mean all the way out to the edges. there should be a backup in the doctor’s office that lets them access medical records when the internet is down.

I don’t care as much for stuff like Netflix because you can always just not watch Netflix, but IMO it’s really really unhealthy for orgs to depend on infinite uptime when other people are depending on them

update: i turned this into a blog post: jyn.dev/brownouts-reveal-syste

Antoine Alberti boosted:
Bradalot “:verified:”bradr@infosec.exchange
2025-11-19

A surprising amount can be read from the data, and for some users, it can be life-threatening.

The entire WhatsApp profile database (including correlatable phone numbers and face photos) leaked.

heise.de/en/news/3-5-Billion-A

#Privacy #WhatsApp #Databreach

Antoine Alberti boosted:

How in the year 2025 are there still places that require a new password every 90 days?!?

Antoine Albertiantoinealberti
2025-11-18

When you come up with a process/framework/whatever, ask yourself at least these questions:
- where in the process do the users smile?
- how can we make that smile bigger and/or more frequent?
This is the essence of agile

Antoine Albertiantoinealberti
2025-11-18

Inside a sprint, the purpose of is to follow the plan, negotiated as a contract at the beginning of the sprint, using processes and tools. At least it talks about working software.

And don't even push me on safe, where the sprint is stretched to 3-month increments. , yeah 🫩

Antoine Alberti boosted:
2025-11-17

Link Post: Behold - the Future of Poltics! smbc-comics.com/comic/conspira

Antoine Alberti boosted:
FediThing 🏳️‍🌈FediThing@chinwag.org
2025-11-16

If AI/LLMs are cut off from stolen human labour, their outputs turn to meaningless gibberish due to model collapse.

If AI/LLM owners are forced to pay for human labour instead of stealing it, they go bankrupt.

If AI/LLM data centres are forced to stay within energy consumption levels compatible with fighting climate change, they have to close down the overwhelming majority of their capacity.

AI/LLMs are just not sustainable, they can only function through labour theft and burning the planet. They're not so much new technology as a new type of ponzi scheme.

Antoine Alberti boosted:
Natasha Jay 🇪🇺Natasha_Jay@tech.lgbt
2025-11-16

Every ad now

🖊️ Cartoon by Stephen Collins

#AI

[Scene is a kitchen - a middle aged woman called JANET is boiling peas at the stove.] 

JANET: Ugh... 

LIZ: What's up? 

JANET: I am so bored of cooking peas! 

LIZ: Have you tried... AI peas?

 JANET: AI peas?

 LIZ: They're peas with AI! [Liz holds up to us a packet of peas labelled: Pea-i AI - Peas with AI]. 

LIZ: Al-powered peas harness the potential of your peas 

JANET: What 

LIZ [Now a voiceover as we cut to a whizzy technology diagram of peas all connected by meaningless dotted lines] Why not take your peas to the next level with Al Peas' new Al tools to power your peas? [Show a techno diagram of a pea with a label reading 'AI' pointing to a random zone in it] 

LIZ: Each pea has Al in a way we haven't quite worked out yet but it's fine [Show Janet and Liz now in a Matrix-style world of peas] 

LIZ: With Al peas you can supercharge productivity and make AI work for your peas! 

JANET: What 

LIZ: Shut up 

LIZ: Our game-changing Pea-Al gives you the freedom to unlock the potential of the power of the future of your peas workflow From opening the bag of peas to boiling the peas to eating the peas To spending millions on adding Al to the peas and then having to work out what that even means. 

JANET: Is it really necessary to- 

LIZ [Grabbing Janet by the collar]: THE PEAS HAVE GOT AI, JANET 

[Cut to an advert ending screen, with the bag of peas and the slogan: AI PEAS: Just 'Peas' for god's sake buy the AI peas. [Ends]
Antoine Alberti boosted:
2025-11-16

@inthehands

I had a conversation with the head of my kid's school about AI

I pointed out that _when an AI is good at a task_ that is usually a _condemnation of the task_ ("why show me a 350-word essay when you could just show me the 40-word prompt instead?")

yours is the same insight, but for music

Antoine Alberti boosted:
2025-11-16

“You are not worse than the kernel hacker, the compiler engineer, or the game engine programmer. You, too, can build whatever you want to build. All software is just software. It's time we treated it that way.”

jimmyhmiller.com/overly-humble

Antoine Alberti boosted:
Allen but one of the good onesuastronomer@mastodon.monoceros.co.za
2025-11-13

Thinking about all these repeated attempts to strip away privacy to make it easier for cops to catch bad guys.

And honestly, I get their point. Figuring out who dun the crime is really hard, there's a reason TV detectives are admired as such super-smart people.

But... I think it was Pratchett who wrote in one of his books that if it were up to the police, everybody in a city would be required to sit still in the open with their hands visible, to make their jobs easier. It would certainly reduce crime!

But the only reason they're asking for this, specifically, is because people don't really understand what strong encryption is really for. If they were to ask for similar access to our physical lives, the push-back would be much, much stronger.

"Criminals often conceal themselves, or 'Hide Out' in residential buildings. They make use of so-called 'Curtains' or 'Drapes' to ensure that they remain hidden. This is why we want all homes to be built without curtain rails, introduce a curtain license (biometric ID required on application), and all curtains are to be fitted with a remote control device allowing any officer to open it from the outside. To alleviate privacy concerns, no officer will be allowed to operate their standard-issue curtain remote control without a warrant."

"Police were unable to capture a wanted criminal last night when they were stymied by a black market door-locking device known as a 'dead-bolt'. These ominously-named devices are designed to secure a door against entry by police officers and anybody else the occupant may wish to exclude. Privacy advocates point out that there are many other legitimate uses for these dead-bolts"

Antoine Alberti boosted:
Nathan Arthurnarthur
2025-11-13

@alexthurow @rjbs debugging is the art of invalidating your cached mental model of how the code works

Antoine Alberti boosted:
Dare Obasanjocarnage4life@mas.to
2025-11-13

The EU seems set to dial back some of its most restrictive tech privacy rules including allowing websites to set some cookies without cookie banners being required and eliminating restrictions on usage of sensitive data such as health information to train AI.

AI comes for us all. Even privacy nerds.

politico.eu/article/brussels-k

Antoine Alberti boosted:
Steve Randy Waldmaninterfluidity@zirk.us
2025-11-13
Text:

I have good news for you, though: assholes are a minority. People of conscience, people with good will and good intentions have always outnumbered psychopaths and sycophants. It might not feel that way but that's because psychopaths have a structural advantage: normal people aren't obsessed with climbing hierarchies and dominating others.

Look around you. The world runs on kindness and empathy. Hardly anything in our daily life would function at all if complete assholes were anything more than a tiny minority of the population. It actually takes a vast machinery to suppress this fact, and to reward evil behaviour.
Antoine Albertiantoinealberti
2025-11-13

We have this stream where one person's the bottleneck. We obviously won't scale them due to, you know, reality, so that's a perfect occasion to experiment with the theory of constraints.
Yeah, scaling is only the last resort in that framework. Interesting 🤔

Antoine Alberti boosted:
2025-11-10

They had me at the headline: AI isn’t replacing jobs. AI spending is

"From Amazon to General Motors to Booz Allen Hamilton, layoffs are being announced and blamed on AI. Amazon said it would cut 14,000 corporate jobs. United Parcel Service (UPS) said it had reduced its management workforce by about 14,000 positions over the past 22 months. And Target said it would cut 1,800 corporate roles. Some academic economists have also chimed in: The St. Louis Federal Reserve found a (weak) correlation between theoretical AI exposure and actual AI adoption in 12 occupational categories."

"Yet we remain skeptical of the claim that AI is responsible for these layoffs. A recent MIT Media Lab study found that 95% of generative AI pilot business projects were failing. Another survey by Atlassian concluded that 96% of businesses “have not seen dramatic improvements in organizational efficiency, innovation, or work quality.” Still another study found that 40% of the business people surveyed have received “AI slop” at work in the last month and that it takes nearly two hours, on average, to fix each instance of slop. In addition, they “no longer trust their AI-enabled peers, find them less creative, and find them less intelligent or capable.”

fastcompany.com/91435192/chatg

Antoine Alberti boosted:
2025-11-10

@dabertime I always find that time works better for making people understand.

A million seconds is 12 days
A billion seconds is 32 years
A trillion seconds is 32,000 years

Antoine Alberti boosted:
Tom Morristommorris
2025-11-10

Microservices adds lots of benefits to the development process like:

- network call overhead for every operation
- more cloud infrastructure to manage—IAM policy writing, costs accounting, etc.—on all those services
- submitting PRs on 50 Git repos in order to make cross-cutting changes
- reimplementing a type system in a YAML file
- buying developer laptops with 64GB of RAM to spin up 50 containers to replicate the functionality previously provided by a damn bash script

Client Info

Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst