Matt TK Taylor's Avatar

Matt TK Taylor

@tk.gg

pm on developer platform @cloudflare.social ex pm @financialtimes.com @MattieTK on the bird app, tk.gg elsewhere. Builder of vidgrid.tk.gg and @project-push.tk.gg

2,659
Followers
1,256
Following
767
Posts
25.04.2023
Joined
Posts Following

Latest posts by Matt TK Taylor @tk.gg

Headless WordPress? Still? 😭

06.03.2026 12:07 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
06.03.2026 12:03 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Lime's billing model is encouraging cyclists to run red lights Per-minute billing on rental bikes may be the worst incentive for good behaviour.

Incredible anecdote.

I've heard similarly from various insiders after writing this interactive graphic that got a lot of attention last month tk.gg/posts/lime-b...

06.03.2026 11:38 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Step 1: Get hired by Lime.
Step 2: Implement stopped bike and junction detection so you don't get charged money to stop at red lights.
Step 3: Quit Lime.

06.03.2026 10:57 πŸ‘ 6 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

To be fair to the AP, that story seemed like baloney. The quotes in it were all in reference to Kalshi buying the AP's election feed rather than any money or service going the other way. It was total spin from what I read of it.

04.03.2026 15:30 πŸ‘ 6 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
Exclusive: It’s bots vs. reporters at the AP The tensions inside the wire service reveal a broader conflict playing out across the media over how AI should be applied within journalism.

AP journalists leaked an internal message from a product manager that annoyed them.

This is quite stunning. Senior PMs typically don't have reports, in news organizations limited power over strategy. Crazy that they'd put her on blast like this.

04.03.2026 10:50 πŸ‘ 5 πŸ” 1 πŸ’¬ 2 πŸ“Œ 0

Also embarrassing to cover because news wires are known to have already experimented in this space because speed is such a critical factor in their product. For the AP newsroom of all people to act like this is new should have been more closely criticised.

04.03.2026 12:09 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

This is great. Can you provide an API that lists committee hearings and their livestreams?

04.03.2026 11:15 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
A post from X head of Product Nikita Bier, known otherwise for bait apps for tweenagers, stating: "Today we are revising our Creator Revenue Sharing policies to maintain authenticity of content on Timeline and prevent manipulation of the program. 

During times of war, it is critical that people have access to authentic information on the ground. With today’s AI technologies, it is trivial to create content that can mislead people.

Starting now, users who post AI-generated videos of an armed conflictβ€”without adding a disclosure that it was made with AIβ€”will be suspended from Creator Revenue Sharing for 90 days. Subsequent violations will result in a permanent suspension from the program.

This will be flagged to us by any post with a Community Note or if the content contains meta data (or other signals) from generative AI tools. 

We will continue to refine our policies and product to ensure X can be trusted during these critical moments."

A post from X head of Product Nikita Bier, known otherwise for bait apps for tweenagers, stating: "Today we are revising our Creator Revenue Sharing policies to maintain authenticity of content on Timeline and prevent manipulation of the program. During times of war, it is critical that people have access to authentic information on the ground. With today’s AI technologies, it is trivial to create content that can mislead people. Starting now, users who post AI-generated videos of an armed conflictβ€”without adding a disclosure that it was made with AIβ€”will be suspended from Creator Revenue Sharing for 90 days. Subsequent violations will result in a permanent suspension from the program. This will be flagged to us by any post with a Community Note or if the content contains meta data (or other signals) from generative AI tools. We will continue to refine our policies and product to ensure X can be trusted during these critical moments."

Almost unbelievabley, X, the everything app, might have made a decision with positive ramifications for the suppression of misinformation.

03.03.2026 17:44 πŸ‘ 4 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
Announcing npmx: a fast, modern browser for the npm registry Today we're releasing the alpha of npmx.dev – a fast, modern browser for the npm registry, built in the open by a growing community.

Really excited for @patak.cat, @danielroe.dev and the entire javascript community for what's going on over at @npmx.dev

This area has needed care and attention for a long time and boy do these guys have care for it.

npmx.dev/blog/alpha-r...

03.03.2026 14:18 πŸ‘ 10 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

ah how frustrating!

03.03.2026 11:34 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Sorry for jumping in: does this oauth spec not allow you to request fewer scopes for an app than it is authorised to request? 🀨

03.03.2026 11:24 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
In fact, Amodei already answered the question: if nuclear weapons were developed by a private company, and that private company sought to dictate terms to the U.S. military, the U.S. would absolutely be incentivized to destroy that company. The reason goes back to the question of international law, North Korea, and the rest:

International law is ultimately a function of power; might makes right.
There are some categories of capabilities β€” like nuclear weapons β€” that are sufficiently powerful to fundamentally affect the U.S.’s freedom of action; we can bomb Iran, but we can’t North Korea.
To the extent that AI is on the level of nuclear weapons β€” or beyond β€” is the extent that Amodei and Anthropic are building a power base that potentially rivals the U.S. military.
Anthropic talks a lot about alignment; this insistence on controlling the U.S. military, however, is fundamentally misaligned with reality. Current AI models are obviously not yet so powerful that they rival the U.S. military; if that is the trajectory, however β€” and no one has been more vocal in arguing for that trajectory than Amodei β€” then it seems to me the choice facing the U.S. is actually quite binary:

Option 1 is that Anthropic accepts a subservient position relative to the U.S. government, and does not seek to retain ultimate decision-making power about how its models are used, instead leaving that to Congress and the President.
Option 2 is that the U.S. government either destroys Anthropic or removes Amodei.

In fact, Amodei already answered the question: if nuclear weapons were developed by a private company, and that private company sought to dictate terms to the U.S. military, the U.S. would absolutely be incentivized to destroy that company. The reason goes back to the question of international law, North Korea, and the rest: International law is ultimately a function of power; might makes right. There are some categories of capabilities β€” like nuclear weapons β€” that are sufficiently powerful to fundamentally affect the U.S.’s freedom of action; we can bomb Iran, but we can’t North Korea. To the extent that AI is on the level of nuclear weapons β€” or beyond β€” is the extent that Amodei and Anthropic are building a power base that potentially rivals the U.S. military. Anthropic talks a lot about alignment; this insistence on controlling the U.S. military, however, is fundamentally misaligned with reality. Current AI models are obviously not yet so powerful that they rival the U.S. military; if that is the trajectory, however β€” and no one has been more vocal in arguing for that trajectory than Amodei β€” then it seems to me the choice facing the U.S. is actually quite binary: Option 1 is that Anthropic accepts a subservient position relative to the U.S. government, and does not seek to retain ultimate decision-making power about how its models are used, instead leaving that to Congress and the President. Option 2 is that the U.S. government either destroys Anthropic or removes Amodei.

Ben Thompson making a full-throated case for fascism here stratechery.com/2026/anthrop...

02.03.2026 17:02 πŸ‘ 468 πŸ” 66 πŸ’¬ 60 πŸ“Œ 49

bah! they must have switched it off, I'll hide it. Sometimes these networks will produce a free live stream during a big event but then revert it to something less interesting afterwards. I suspect that's what's happened here.

01.03.2026 12:27 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
Cirrus A very easy, single-user Bluesky/ATProto Personal Data Server

You could look at Cirrus from my very smart colleague @mk.gg mk.gg/projects/cir...

R2 is going to cost you approximately $3 a month at that capacity

01.03.2026 06:35 πŸ‘ 4 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

"Get the coordinates we need for the strike. Make no mistakes."

01.03.2026 06:19 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

btw @cloudflare.social gives you 10hrs of remote browser time a month with the $5 plan, that's easily 3000 unique urls with unique time-sensitive OG images

28.02.2026 18:32 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

wassuppp turns out all I really needed to replace was the puppeteer workers binding (powerful, slow) with the rest api (fast, not a binding, requires an api key, boo).

bsky.app/profile/tk.g...

28.02.2026 18:24 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
VidGrid - watch multiple news channels in your browser A free news focused multiviewer for editors and journalists

omg it kinda works (and I've shipped it so whatever)

this is a dynamic opengraph image produced through a remote browser rendering the pasted url on cloudflare's network, taking ~8s to produce.

Really. Freaking. Fun. Internet. Shit.

vidgrid.tk.gg?layout=N4IgN...

28.02.2026 18:23 πŸ‘ 4 πŸ” 0 πŸ’¬ 3 πŸ“Œ 1

Right, it's bedtime here in GMT+8.

I have a 3hr flight tomorrow so everyone let me know what your favourite emergency pod was.

28.02.2026 17:22 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

think I might need to get rid of nextjs and replace it with a lightweight alternative πŸ€”πŸ€”πŸ€”πŸ€”πŸ€”πŸ€”πŸ€”πŸ€”πŸ€”πŸ€”

28.02.2026 16:59 πŸ‘ 3 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

mate claude did that aeons (months) ago

i'm overengineering an /og endpoint to use browser rendering to produce dynamic OG share cards but currently its a bit slow....

28.02.2026 16:58 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I'm afraid I'm far too busy implementing something infinitely more useless

28.02.2026 15:52 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Ooh, could you have this set of labels etc powered by a GitHub repo and GitHub actions? I'm sure lots of people would be open to helping you out with the orgs, labelling, and descriptions.

28.02.2026 14:58 πŸ‘ 3 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
In Trump’s Case for War, a Series of False or Unproven Claims

NEW: The Trump administration's 3 key claims about Iran's nuclear & missile programs are false or unproven β€” including Iran "soon" making a missile that can hit the US. Intel analysts are concerned about aides exaggerating threats and misusing intelligence. Gift link: www.nytimes.com/2026/02/26/u...

27.02.2026 02:37 πŸ‘ 208 πŸ” 99 πŸ’¬ 19 πŸ“Œ 13

thanks for sharing @eliothiggins.bsky.social πŸ₯°

28.02.2026 14:42 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Post image

Internet connectivity in Iran is now in a near-complete shutdown, down about 98% vs the previous week. The disruption began around 07:00 UTC, following reports of military strikes β€” and in past incidents the government has responded with nationwide Internet shutdowns.

28.02.2026 12:03 πŸ‘ 816 πŸ” 299 πŸ’¬ 40 πŸ“Œ 13
Preview
Becoming bankrupt How to apply for bankruptcy. What to do if you’ve been made bankrupt: if you’re at risk of violence, how long it lasts, bankruptcy restrictions, what happens to your home and possessions when you're b...

opening this url without a tv license redirects you after 30 seconds to www.gov.uk/becoming-ban...

28.02.2026 10:34 πŸ‘ 3 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0
Post image

BBC/NBC/AJE/Sky all in one grid with one click audio switching for those following US/Israel/Iran activity this morning

vidgrid.tk.gg?layout=N4IgN...

28.02.2026 10:04 πŸ‘ 11 πŸ” 5 πŸ’¬ 1 πŸ“Œ 0