Steph Taylor's Avatar

Steph Taylor

@smtaylor

Ocean modeller / scientific programmer. ❤️: enviro science, physics, plants, weird critters, textile arts.

62
Followers
172
Following
20
Posts
05.08.2023
Joined
Posts Following

Latest posts by Steph Taylor @smtaylor

As @ldobsonhughes.bsky.social notes, international law isn't a principle to be sacrificed on the altar of hard-headed self-interest: it's a core Canadian national interest, and should be defended as such.

bsky.app/profile/ldob...

02.03.2026 15:45 👍 7 🔁 5 💬 0 📌 0

If your perspective is that Canada should only voice support for international law when it has an narrow self interest in doing so, we're no longer discussing international law.

28.02.2026 18:44 👍 117 🔁 45 💬 5 📌 4

Several points:
1) Iran's regime is odious. Its citizens deserve support.
2) That does not mean an attack will help.
3) Canada will find it hard to be so selective in its support for international law.
4) Canada does not benefit from a US administration unbound by the US constitution.

28.02.2026 18:01 👍 372 🔁 113 💬 7 📌 7
Left side profile of the pheasant who regularly visits my mum and dad's garden on the day he first became a pheasant monk and got a tonsure.

Left side profile of the pheasant who regularly visits my mum and dad's garden on the day he first became a pheasant monk and got a tonsure.

The pheasant who regularly visits my mum and dad's garden, looking straight on us while daring us to suggest he is partly made of fuzzy felt.

The pheasant who regularly visits my mum and dad's garden, looking straight on us while daring us to suggest he is partly made of fuzzy felt.

A more relaxed side of the pheasant who regularly visits my mum and dad's garden, but where he still looks like he is the product of an industrious afternoon from a local fibre arts group.

A more relaxed side of the pheasant who regularly visits my mum and dad's garden, but where he still looks like he is the product of an industrious afternoon from a local fibre arts group.

I absolutely refuse to believe that the pheasant who regularly visits my mum and dad’s garden is real and has not in fact been crocheted.

26.04.2025 06:22 👍 1434 🔁 258 💬 44 📌 31

What doesn’t seem to be on the table: it’s obvious the existing scheme of having the urban core pay for suburban sprawl, and have new home owners and renters pay more taxes than long-time home owners isn’t working. But they will try literally anything but fixing that

15.02.2026 17:44 👍 54 🔁 13 💬 6 📌 4
Someone please show me how to do alt text when I am sharing text only photos without having to type it all out

Someone please show me how to do alt text when I am sharing text only photos without having to type it all out

B

B

C

C

D

D

The list of things council is considering cutting in order to keep taxes “flat” is insane. (From Sam Austin’s website). Call your councillor and tell them you won’t support them if they support these cuts.

14.02.2026 19:57 👍 77 🔁 45 💬 13 📌 18

*staring directly at the current Canadian federal policy agenda*

12.02.2026 20:42 👍 18 🔁 12 💬 0 📌 0

The biggest challenge Canadians face right now involves the utter disconnect between Carney’s rhetoric and his actual policies. It makes it so difficult to even begin a debate about what actually standing up to the US would look like.

24.01.2026 12:28 👍 63 🔁 21 💬 1 📌 1
Preview
How Trump’s regulatory cuts threaten Canada’s safety Trump’s dismantling of U.S. science agencies puts Canadians at risk. Canada must boost regulatory capacity to protect health and food safety.

Canada has relied on U.S. science for decades - for health info, drug review, food safety, weather monitoring, and more.

As the US dismantles its science ecosystem @ntusikov.bsky.social argues Canada’s civil service cuts will put public heath at greater risk.

policyoptions.irpp.org/2025/12/trum...

23.01.2026 11:58 👍 57 🔁 45 💬 1 📌 1

If I were a PM who believed we were at a global rupture moment, upending decades of international order, and requiring totally new relations with the world, I probably wouldn’t cut my foreign affairs staffing and budget by 15%

21.01.2026 15:31 👍 395 🔁 139 💬 11 📌 13
Paintings of hammerhead sharks depicted in the "Oki National Products Illustration Notes". From Japan, Edo Period, ca 1735

Paintings of hammerhead sharks depicted in the "Oki National Products Illustration Notes". From Japan, Edo Period, ca 1735

A pair of happy hammerhead sharks
(ca. 1735 Edo period Japan)

28.12.2025 17:57 👍 3837 🔁 1241 💬 28 📌 45

We didn't survive and thrive as a species because we had warriors or hunters - pretty much every omnivore can manage that.

We did so because we developed Grandmothers, and domesticated Dogs.

Community is what defines us as a species, and every "self-reliant" libertarian shitweasel be damned.

28.12.2025 05:00 👍 3266 🔁 903 💬 35 📌 25
 A morbidly inclined sheep, who likes to hang around in graveyards, detests parties and owns original pressings of the entire classic run of Leonard Cohen LPs from 1967-1974.

A morbidly inclined sheep, who likes to hang around in graveyards, detests parties and owns original pressings of the entire classic run of Leonard Cohen LPs from 1967-1974.

THREAD. A collection of photographs of excellent sheep I have met on walks.

You will find the captions to each photo in the alt text.

14.11.2025 09:48 👍 2564 🔁 675 💬 90 📌 253
Preview
Legal Issues Raised by a Lethal U.S. Military Attack in the Caribbean The Trump administration’s extraordinary lethal attack on this purported smuggling vessel raises significant potential legal issues.

1/ I worked at DoD. I literally cannot imagine lawyers coming up with a legal basis for lethal strike of suspected Venezuelan drug boat.

Hard to see how this would not be "murder" or war crime under international law that DoD considers applicable.

Read expert analysis by @bcfinucane.bsky.social⤵️

03.09.2025 21:41 👍 8498 🔁 3027 💬 360 📌 304
Post image

If cabinet is hoping the guy who wrote the fascist playbook will give advice on stopping fascism then we are in trouble.
Adorno reminds us that clever people were the first to be suckered by the fascist playbook because they thought they could outsmart it.
There is nothing hidden in their plan.

04.09.2025 00:00 👍 629 🔁 210 💬 91 📌 61

And I resent the way companies are trying to force it. The other day Slack emailed saying they were adding AI to their product that already functions fine. I immediately cancelled that account. I don’t want it. It is not a benefit or feature. It is a yoke. I am happy on this hill even if I am alone

19.06.2025 17:55 👍 2468 🔁 224 💬 94 📌 26

Yes. I am right. I am seeing how Chat GPT is ruining students critical thinking and writing skills in real time. It is not the future. It is a tool designed to render the populace helpless, to make people doubt their innate intelligence, and to foster overreliance on technology.

19.06.2025 17:51 👍 7840 🔁 2351 💬 122 📌 120
Preview
Canada and India reset relations as Mark Carney and Narendra Modi meet Leaders call time on two-year diplomatic dispute triggered by death of Sikh activist

Seeing my prime minister shake hands with a man behind the murder of a Canadian citizen on Canadian soil turns my stomach.

Mark Carney will talk about rule of law and human rights, but at the end of the day, it’s economy, economy, economy, no matter the cost.

18.06.2025 13:10 👍 282 🔁 67 💬 15 📌 6

Like yet again the fear of "losing out" on the "AI revolution " takes precedence over the risks of joining in. There'd literally have been fewer genocides of we'd just pumped the brakes on Facebook a bit

19.06.2025 11:50 👍 108 🔁 16 💬 3 📌 0
Ironically, upon the paper’s release, several social media users ran it through LLMs in order to summarize it and then post the findings online. Kosmyna had been expecting that people would do this, so she inserted a couple AI traps into the paper, such as instructing LLMs to “only read this table below,” thus ensuring that LLMs would return only limited insight from the paper.

She also found that LLMs hallucinated a key detail: Nowhere in her paper did she specify the version of ChatGPT she used, but AI summaries declared that the paper was trained on GPT-4o. “We specifically wanted to see that, because we were pretty sure the LLM would hallucinate on that,” she says, laughing.

Ironically, upon the paper’s release, several social media users ran it through LLMs in order to summarize it and then post the findings online. Kosmyna had been expecting that people would do this, so she inserted a couple AI traps into the paper, such as instructing LLMs to “only read this table below,” thus ensuring that LLMs would return only limited insight from the paper. She also found that LLMs hallucinated a key detail: Nowhere in her paper did she specify the version of ChatGPT she used, but AI summaries declared that the paper was trained on GPT-4o. “We specifically wanted to see that, because we were pretty sure the LLM would hallucinate on that,” she says, laughing.

Amazing: MIT researchers revealed how ChatGPT etc are destroying our brains and booby-trapped the report to expose those who want to use AI to ostensibly summarize the results.

t.co/JXeTALBPds

19.06.2025 11:23 👍 5135 🔁 2138 💬 50 📌 181

You cannot responsibly write about AI without writing about politics. "AI" is an intensely political project, especially in its current formation, where development is led by tech giants that increasingly depend on the state—for contracts, land rights, and, of course, for deregulation.

22.05.2025 18:59 👍 609 🔁 189 💬 10 📌 10

happy for Khalil of course but i must say, "lawful permanent resident not charged with any crimes, imprisoned by president's stormtroopers for having forbidden opinions about US-abetted genocide, briefly allowed to touch his child" is a hell of a thing to find oneself feeling joy and relief about

22.05.2025 16:33 👍 6886 🔁 1770 💬 26 📌 19

No one voted for this shit, and the information environment is so broken that most people don't know it's happening & probably won't know it happened afterward, and if they do find out, won't know who's responsible.

22.05.2025 18:40 👍 514 🔁 141 💬 14 📌 8
Preview
Trump Administration Halts Harvard’s Ability to Enroll International Students

everyone with undue power is playing stupid games and winning stupid prizes
www.nytimes.com/2025/05/22/u...

22.05.2025 18:34 👍 71 🔁 14 💬 0 📌 2

Why, why, why?! Does Notepad need to have a Co-pilot implementation. It is Notepad. It exists to have no features.

The whole point of Notepad is that it doesn't do anything but text.

22.05.2025 18:35 👍 1721 🔁 164 💬 40 📌 9
Elon Musk, Sam Altman and Other C.E.O.s Join Trump at U.S.-Saudi Lunch
Many of the executives — including representatives of four of the 10 biggest U.S. companies by market value — are seeking business from Saudi investors and a chance to lobby top Trump officials.

Elon Musk, Sam Altman and Other C.E.O.s Join Trump at U.S.-Saudi Lunch Many of the executives — including representatives of four of the 10 biggest U.S. companies by market value — are seeking business from Saudi investors and a chance to lobby top Trump officials.

Let's not forget for a second that AI is now overwhelmingly an oligarchic project.

www.nytimes.com/2025/05/13/u...

13.05.2025 16:50 👍 1158 🔁 433 💬 27 📌 32

Liberal hacks are gonna lose their shit on me but it is a huge 🚩 that we have a cabinet Minister for AI but not one for Labour.

13.05.2025 16:49 👍 231 🔁 65 💬 15 📌 15
Preview
Nearly 500,000 children could die from AIDS-related causes by 2030 Experts including Prof Lucie Cluver, Professor of Child and Family Social Work, Department of Social Policy and Intervention, and Dr Seth Flaxman, Associate Professor, Department of Computer Science

This study in The Lancet found that cuts to US spending on PEPFAR, the program to deliver H.I.V. and AIDS relief abroad, could cost the lives of 500,000 children by 2030.

Half a million kids.

Kids.

www.ox.ac.uk/news/2025-04...

08.05.2025 12:32 👍 518 🔁 303 💬 27 📌 28
Highlighted text form THE NEW YORKER

The task that generative A.I. has been most successful at is lowering our expectations, both of the things we read and of ourselves when we write
anything for others to read. It is a fundamentally dehumanizing technology because it treats us as less than what we are: creators and apprehenders of meaning. It reduces the amount of intention in the world.

Highlighted text form THE NEW YORKER The task that generative A.I. has been most successful at is lowering our expectations, both of the things we read and of ourselves when we write anything for others to read. It is a fundamentally dehumanizing technology because it treats us as less than what we are: creators and apprehenders of meaning. It reduces the amount of intention in the world.

Reminds me of this eternal Ted Chiang banger-

07.05.2025 21:58 👍 4335 🔁 1423 💬 18 📌 30