RE: mastodon.world/@somecanuckchic…
The worst part is that Frankie Legs is staying on until a successor is found...
We have a Deputy Premiere, right? Pass it over to Geneviève Guilbault and let her shit the bed until a successor is chosen.
#polQC #QCpoli #assnat #CAQASTROPHE
somecanuckchick (@somecanuckchick@mastodon.world)
Frankie Legs talking about how he'd taken the last month to think about the future of the CAQ and his role in the CAQ... I saw him at a Habs game in Montréal...somecanuckchick (Mastodon)


Dmytri
in reply to Dmytri • • •2/9
Dmytri
in reply to Dmytri • • •3/9
Dmytri
in reply to Dmytri • • •4/9
Dmytri
in reply to Dmytri • • •With spec-driven workflows, agile tight inner loops become slow outer loops that depend on high-fidelity specifications.
Work shifts from coding to making specs, prompts, lemmas, tests, etc. The agent generates code really fast, but feedback becomes hidden.
5/9
Dmytri
in reply to Dmytri • • •Debugging moves from fixing code to adjusting specs and tests.
You must anticipate correctness conditions in advance, and creating the specifications and verification is also technical, sometimes even more so than coding.
6/9
Dmytri
in reply to Dmytri • • •7/9
Dmytri
in reply to Dmytri • • •Over time, the codebase becomes Ptolemaic: as the agent adds epicycles, guards, exceptions, and compensations to keep the system passing its proofs.
The result is a Pyrrhic victory: The time saved writing code is instead consumed by specification and verification.
8/9
Dmytri
in reply to Dmytri • • •9/9
Duchamp Pérez
in reply to Dmytri • • •One thing that XP got right is that often the specs emerge from the feedback loop. We are learning more about the business problem and the technology limitations via coding.
A solution not working is known. The engineer's value is knowing why it didn't work and communicating this to the spec authors
Christina Jennifer
in reply to Dmytri • • •There is a word for the process of creating a specification that is so robust that it can't fail to meet the spec.
Programming. It's called programming.
All you've done is moved to a higher level with a non-deterministic compiler.
Dmytri reshared this.
sw
in reply to Christina Jennifer • • •Dmytri
in reply to sw • • •sw
in reply to Dmytri • • •But not the whole team consist of developers that can code...
But maybe I misunderstood her.
Dmytri
in reply to sw • • •sw
in reply to Dmytri • • •@criffer ah, we speak about AI tools to create code...
Your "spec" must be as good & as exact as possible.
We - in my area - have iterations/sprints/feedback cycles in order to close gaps, clarify misunderstandings, learn, etc with the target to have the product that you want - interaction between e.g. PO and Dev. For me that is one aspect of agile, scrum - speak to each other.
How such a communication can be executed between POs and AI-tools I cannot comment - not my working mode.
Dmytri
in reply to sw • • •Nickynah
in reply to Dmytri • • •The issue I see is that the prospect of “no coding” is so enticing to the industry that no amount of issues will disuade them.
Ignacy 🇪🇺
in reply to Dmytri • • •Daniël Franke
in reply to Dmytri • • •Stephan Lukasczyk
in reply to Daniël Franke • • •Daniël Franke
in reply to Stephan Lukasczyk • • •metr.org/blog/2025-07-10-early…
It's a bit old now, but there has been no study since that wasn't based on vibes only.
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
METR (metr.org)Stephan Lukasczyk
in reply to Daniël Franke • • •Daniël Franke
in reply to Stephan Lukasczyk • • •Stephan Lukasczyk
in reply to Daniël Franke • • •Dmytri
in reply to Stephan Lukasczyk • • •Daniël Franke
in reply to Dmytri • • •I'm the opposite side of that, when I tried LLMs in coding, it just made my job worse and weird bugs kept sneaking in, of the type that would not happen if I wrote it myself. Also, code review of something that's inherently incapable of learning is a special kind of torture.
And if one ends up writing so much boilerplate, it's a code smell signalling that you need to move stuff to a library.
That said, my main problems with LLMs are ethical. Even if it did its job well, I would not use them because of all the externalities.
Dmytri
in reply to Daniël Franke • • •Dmytri
in reply to Daniël Franke • • •Michael Vilain
in reply to Dmytri • • •Dr Susi Arnott
in reply to Dmytri • • •Like getting a good grade by paying someone else to take an exam for you
Rocinante
in reply to Dmytri • • •