Lightnews — Scholar-powered news

LightNews

Chris Cook

@chriscook.news

15K followers 460 following 300 posts

ft senior reporter. south london nationalist. articles with my byline get posted automagically on @cook.report

Posts Media Videos Starter Packs

Chris Cook @chriscook.news · 6d

Not a lot of people familiar with the “political uniforms” legislation

Chris Cook @chriscook.news · 7d

this rule did lead to this delight (he’s been booked for it before) m.facebook.com/story.php?st...

Log in or sign up to view

See posts, photos and more on Facebook.

Chris Cook @chriscook.news · 9d

Hahaha - okay fair enough

Chris Cook @chriscook.news · 9d

pah. While you’re on a history tour, why not include Preston North End and Ipswich

Chris Cook @chriscook.news · 9d

excuse me?

a soccer player wearing a yellow jersey that says netbet

ALT: a soccer player wearing a yellow jersey that says netbet

media.tenor.com

Reposted by Chris Cook

Kaye Wiggins @kayewiggins.ft.com · 12d

As Trump pressures the Department of Justice to target his opponents, my colleagues @chriscook.news, Paul Caruana Galizia and I dug deep into a DoJ unit - the US Attorney’s Office for the Southern District of New York - which had been known for its independence.

Full story: on.ft.com/42VG5IK

Rule of law in the Trump era: the demoralised prosecutors of SDNY

[FREE TO READ] The Southern District of New York once took pride in its independence. But pressure from the administration is changing the way its lawyers operate

Chris Cook @chriscook.news · 12d

people need to lay off Ruben Amorim — I really think he’s got it in him to manage a top 6 finish and coming back up via the playoffs next year

Chris Cook @chriscook.news · 14d

We should give Jim Ratcliffe another knighthood for contribution to gaiety of the nation.

Reposted by Chris Cook

Helen Warrell @helenwarrell.bsky.social · 15d

What is Russia's spy ship Yantar and why has it been loitering in British waters? The FT investigates the threat to undersea cables & the work of Moscow's secret deep-sea research unit, GUGI, as an agent of sabotage. With @chriscook.news Daria Mosolova & David Djambazov

www.ft.com/content/0b35...

The Russian spy ship stalking Europe’s subsea cables

Covert operations in waters surrounding the British Isles pose a grave threat to critical infrastructure and a fresh challenge to Nato

Chris Cook @chriscook.news · 15d

The undersea war is a major problem for the UK and Ireland, and one both countries will need to take seriously

cookbot 🤖 @cook.report · 15d

🚨 New article by @chriscook.news

The Russian spy ship stalking Europe’s subsea cables

Covert operations in waters surrounding the British Isles pose a grave threat to critical infrastructure and a fresh challenge to Nato

Chris Cook @chriscook.news · 16d

@davidgerard.co.uk given that your allegations were wrong, can you please delete your posts about us, which are still online for some reason and free for other people to see and share.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is just wrong bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

Just to be clear, that guy is completely 100% wrong. bsky.app/profile/chri...

Chris Cook @chriscook.news · 17d

This is wrong.

a) The article literally states we did the analysis with AI. We used semantic search to ID paragraphs relating to a topic, then grouped text into categories by their meaning.

b) The model outputs were, as below, boolean indicators about whether something fit a category.

Chris Cook @chriscook.news · 17d

no, it did something we could never have done -- reviewing and sorting 7,500 separate paragraphs across five years of 10-Ks/10-Qs and one year of earnings calls. it created work for me, it took it from no-one.

Chris Cook @chriscook.news · 17d

they paid me to research, write and edit this. we did this. the ai did none of those things.

Chris Cook @chriscook.news · 17d

I think it's three quotes now?

Chris Cook @chriscook.news · 17d

i have a lot on this week. but using AI tools to group them by their own meanings (without regard to anything else-- just reading the text) is something we do a lot. you don;t need a load of the old logits or ML stuff we used for this. it's better, and faster -- as we know from validation tests

Chris Cook @chriscook.news · 17d

i'd like to make a correction to my correction: actually it's fine, i misread my own copy.

Chris Cook @chriscook.news · 17d

we did! we used the API to categorise quotes! "is this text in a well defined category" is a really good use of them! so you can ask them "does this mention XYZ" and get them to reply TRUE/FALSE. Then we do validation exercises to check they do what we wanted. The *TEXT* never comes out of it.