Tootfinder

Friday Links 26-32
The trend is definitely going towards a smaller Engineering and a bigger AI section.
I don't like that.
This week, have a look at John Cutler's Feature Factory Retrospective and the No Man Is an Island post.
If you want a podcast and like engineering, check out "Building resilient systems".

TSMC says it plans to build an additional 3nm fab in Tainan, Taiwan, set for production in H1 2028, alongside its 3nm capacity in Arizona and Japan's Kumamoto (Qianer Liu/The Information)
https://www.theinformation.com/briefings/tsmc-add-new-3nm-fab-ai-orders…

TSMC to Add a New 3nm Fab as AI Orders Pile up
Taiwan Semiconductor Manufacturing Co. said Thursday it will build another 3-nanometer chip factory in the Taiwanese city of Tainan, with production set for the first half of 2028. The move signals confidence that strong AI chip demand will last for years. CEO C.C. Wei said strong customer demand prompted the decision to build the new 3nm factory in addition to other planned 3nm plants in

Things you don't notice until you actually get the instrumentation to study it: some of my smaller IT gear has a terrible power factor.
The average power factor of my main IT rack lower PDU (running most of the servers and switches) is an excellent 0.99 with about 1 kW of load.
The upper PDU on the main rack is averaging 0.86 with about 350W load, with the main offenders being the FS 1G edge switch (14W @ 0.58 PF) and the PoE injector (48W @ 0.57 PF).
But the PDU on the…

So #Russia just lost the capacity to put stuff in space, apparently. Either immediately, or in the very near future.
Ukraine hit its main rocket factory in Samara.
In part, it's a shame. But it was inevitable, as the very same rockets that sent stuff to the ISS are putting military crap in orbit. So this factory is an obvious military target.

#Gewalt im #Alter
https://www.suedtirolnews.it/wirtschaft/wenn-fuersorge-zur-gewal…

Wenn Fürsorge zur Gewalt wird
Bozen – Gewalt hat viele Formen, die mit unterschiedlichsten Faktoren zusammenhängen können. Auch das Alter kann dabei ein Faktor sein. In einer gemeinsamen Sensibilisierungsaktion wollen darum das Land Südtirol und die Gemeinde Bozen auf Formen von Gewalt an Seniorinnen und Senioren hinweisen und aufzeigen, welche Maßnahmen dagegen ergriffen werden können. Dafür wurde eine Broschüre erarbeitet, […]

Don't yell at your colleagues; yell in your shell
https://refp.se/articles/your-shell-and-the-lazy-exclamation-mark

For Indian communities, fishmeal factories bring pollution, danger and protests
Your salmon and prawns depend on this industry. It's bonkers to catch fish, turn them into fishmeal to feed other fish. Don't eat farmed salmon, in particular.
https://

For Indian communities, fishmeal factories bring pollution, danger and protests
A proposed fishmeal plant in Cuncolim on India’s southwestern coast has reignited opposition to an industry accused of pollution, unsafe working conditions, and unsustainable fishing practices. An Outlaw Ocean Project investigation found at least 30 Indian fishmeal plants linked to environmental violations or other legal infractions between 2018 and 2024, as well as 18 episodes of civil unrest. The industry provides jobs and tax revenue, but also diverts fish from local food supplies, exposes…

How Congressional lawmakers and aides are using AI tools with little oversight to write speeches and news releases, sort constituent mail, and more (Anna Liss-Roy/Washington Post)

Chatbots are doing the work of Congress with little oversight
From writing speeches to sorting constituent mail, AI is spreading through Congress faster than the rules governing its use.

Anthropic's text watermark alters word probabilities to embed a fingerprint, which could degrade Claude's writing, despite its claim of no impact on quality (John Gruber/Daring Fireball)
https://daringfireball.net/2026/08/ant

Anthropic’s ‘Watermark’ Text Adulteration in Claude Is a Perversion of Writing
It’s unacceptable for a tool to sacrifice an iota of clarity, coherence, meaning, quality, etc. for the purpose of embedding hidden clues within the text to suggest its provenance. The idea that anything other than *my* needs should factor into the generation of text *for me* is patently offensive.

OpenAI previews Ultrafast, an API tier powered by Cerebras that runs GPT-5.6 Sol up to 14× faster and generates up to 750 output tokens per second (Zac Hall/9to5Mac)
https://9to5mac.com/2026/08/13/openai-previews-ultrafast-gpt-5-6-sol-runni…

OpenAI previews 'Ultrafast' GPT-5.6 Sol running up to 14 times faster - 9to5Mac
OpenAI is previewing a new way to run its most capable GPT-5.6 model at dramatically higher speeds. The company says...