Every second Sunday #FutureHistories #Podcast S02E41 - Antoinette Rouvroy on Algorithmic Governmentality

#AntoinetteRouvroy, #JanGroos, #Interview, #algorithmicGovernmentality, #algorithms, #governmentality, #digitalization, #DasRegierenDerAlgorithmen, #AI, #ArtificialIntelligence, #algorithmicTransformation, #automatization, #data, #bigdata, #knowledge, #future, #emanzipation

https://futurehistories.podbean.com/e/s02e41-antoinette-rouvroy-on-algorithmic-governmentality/

and its official - Marvins first paper „Investigating the effect of intelligent assistance systems on motivational work characteristics in assembly" is published - you can check it out (open access) in the Journal of Intelligent Manufacturing: https://link.springer.com/article/10.1007/s10845-023-02086-4
#proudsupervisor #PhD #iopsychology #workmotivation #automatization #digitalisation #futureofwork

When you do requirements management (in aerospace, automobile, IT, etc.), you implement this routine: 「discuss ⟶ define changes ⟶ change/add a requirement ⟶ write the reason」. Then you cross‑link the requirements to see the effects on your design.

Some software companies try to automate/innovate this process (DOORS, Flow Engineering, MS Office, Valispace, etc.) & solve the main issues: traceability, repairability, fewer mistakes, less human involvement.

Well, that is a wrong process.

No, scratch down. The process is fine — it’s almost ideal & time‑proven: humanity built a lot with it, almost everything is built using it.

But the modern implementations are wrong.

This process is old — e.g., it was implemented in space sectors of every country since the very 1st day of the space era in the 1950s (~70 y ago). But they did it using paper because there was a lack of CPUs. Someone still uses papers company‑wide — a lot of papers/PDFs. Most IT companies try to implement this old routine using keyboards, screens, cloud computing. Nothing new.

Back there, there were limitations of the paper, which are carefully implemented in modern tools. Among the reasons for these limitations: a lot of time/paper required to write down every opinion/argument, lack of space to store the papers, lack of automatization to navigate between them. So, there were people & paper: you write discussions with your hands. Like you do now using a keyboard.

The major issue is in weeks/months you’ll forget why exactly it was done, who discussed it, who made the decision, what was the environment, what led to this decision, etc. Because of the human factor. People are lazy, make mistakes, have problems with communication. Pretty soon everyone tends to pay less attention to the words they write, especially during a rush. No matter what tool you use (one of the mentioned above, an issue tracker, Excel, email, etc.) you’ll have these reasons for changes: 「as per discussed」, 「we change the models」, 「changes in source data」, etc.

So, the main issues are not solved: you don’t understand where is a description of decisions, you write down everything manually, you write what you think is correct. Paying more attention will not work; you can reduce human errors only by reducing the presence of humans. Ideally, requirements management shall not include humans.

The solution is easy. You already have instruments (chats, email, minutes/videos from meetings, etc.) to collaborate with people & decide what to do next — those are tools for 「how, when, who」. You can put a mark in those to note that at this very moment, a decision was made. Let robots arrange requirements & design analysis while putting this moment as a reason for changes. If you’ll want to find out the reasons in several months/years — just click & navigate to those chats/meetings back there.

Easy.

Yet, there is no implemented solution.

Tags: #automatization #engineering #management #requirements

Here is a piece I wrote some years back on how in the #GigEconomy #labor relations are made "technical" through #automatization and making them part of the #algorithms

This "making technical" obfuscates and masks the very real relations of power at the workplace.

Relations that affect what workers, in this case #couriers, do and how much they are paid, are made features of the software, making it harder for the worker to question them.

#Justice4Couriers
https://www.justice4couriers.fi/2018/10/11/when-mr-robot-is-your-boss-working-under-algorithms/

Amigos, les cuento que vi robots llevando el mercado en Leeds.
Adjunto fotos. (Al parecer hab sido un éxito).
#automatization

:catPOWER: Napkin Ideas Around What Changes to Expect Post-ChatGPT

via Daniel Miessler
#ChatGPT #GPT3 #Automatization #Inequality
https://danielmiessler.com/blog/ideas-changes-expect-post-chatgpt/?utm_source=pocket_mylist

:catPOWER: Napkin Ideas Around What Changes to Expect Post-ChatGPT
— Daniel Miessler
#ChatGPT #GPT3 #Automatization #Inequality
https://danielmiessler.com/blog/ideas-changes-expect-post-chatgpt/?utm_source=pocket_mylist

If all visions of the future are essentially about technology, you start to wonder: What relevance would I still have for those who hold control over these technologies?

#harari #automatization #irrelevance

#JavaCodeGeeks #Blog "How and Why to Analyze, Generate and Transform Java Code Using Spoon" #Java #Spoon #framework #library #analyse #generate #transform #Automatization ... https://www.javacodegeeks.com/2019/03/analyze-generate-transform-java-spoon.html

#JavaCodeGeeks #Blog "How and Why to Analyze, Generate and Transform Java Code Using Spoon" #Java #Spoon #framework #library #analyse #generate #transform #Automatization ... https://www.javacodegeeks.com/2019/03/analyze-generate-transform-java-spoon.html

#DZone #JavaZone "JGit Library Examples in Java" #Java #JGit #Git #automatization #library ... https://dzone.com/articles/jgit-library-examples-in-java

#DZone #JavaZone "JGit Library Examples in Java" #Java #JGit #Git #automatization #library ... https://dzone.com/articles/jgit-library-examples-in-java