Thread by @RickyRawls

I'm going to build on this quoted thread and get a little conspiratorial: I don't know if a lot of people know this but Buzzfeed actually started out as a chatbot itself that curated and compiled news stories.

https://twitter.com/RickyRawls/status/1608329358871244804

It was just a chatbot with no humans but the tech wasn't there yet so founder Perretti started hiring humans as editors and content creators to build around the chat bot and create original content

I think they had to use people because the AI and machine learning tech wasn't quite there yet, but from the beginning Buzzfeed ceaselessly has been doing R&D into data science and how it intersects with journalism and creativity starting with its OpenLab

Perretti in interviews has always been open about how they were constantly doing data science research in house to figure out how it could replace humans, although he made sure to downplay the threat it would pose to the writing side of things

They had been working on OpenLab and recruiting creatives and data science people for it for years, although it recently was closed buzzfeed.com/amandahickman/…

Machine learning has always been a major obsession of Buzzfeed and it comes up often in OpenLab job postings. I think the creatives were mainly needed to help mold the machine learning AI as input and modeling

When you read Buzzfeed's investor relations materials there is a constant focus on the promise of how machine learning and AI can be used to help content creators and automate a lot of the decisionmaking in when and what to post. This is what the future they pitch to investors

This is from their 2021 investor relations material: investors.buzzfeed.com/static-files/d…

As you keep moving forward through the years of thier investor materials, this is from 2022 investors.buzzfeed.com/static-files/4…, they get more and more overt on how they are working on AI and machine learning to play a bigger role in the actual content creation

Buzzfeed closed down its OpenLab but has been hiring data science and developing proprietary machine learning data as well as acquiring whole data science companies for long time before and also after OpenLab project. It's always been a major focus; again, they had AI before ppl

For years people have been pointing out how Buzzfeed has been dumbing down journalism and the tone and craft of it. They had a small serious wing of Buzzfeed news that I think was there to do what AI never could, long, difficult investigative journalism, but really it...

...was to help distract from the bigger project of Buzzfeed which was to trivialize and dumb down journalism and the state of media content at 95% of the rest of the site. Like how record labels have loss leader jazz imprints for prestige and social capital but the pop subsidizes

Anyway, pretense of even that proved too unprofitable to maintain, and they got rid of Buzzfeed news and have increasingly gutted more + more creative and editorial but keeping the fluff and growing the data science side. The word data appears over 100 times in investor relations

People have been warning about the "Buzzfeedication" of journalism and blogs and social media for a long time. First article I remember talking about it was @tomscocca's "On Smarm" gawker.com/on-smarm-14765…

What I wonder is this: what if smarm and dumbed down rote childish formulaic writing needed to become the norm because that was what was found as easier for AI to replicate? What if Buzzfeed had to make this the dominant indsutry form of writing to meet the AI halfway?

Everything is Buzzfeedy now. Even the NY times and Washington post has horrible Buzzfeedy type writing now. An article like this for example would never have made NY Times before Buzzfeed normalized it for adults nytimes.com/2021/05/07/mov…

Not only has Buzzfeed been gutting more and more of it's human creative, which it only ever acquired in the first place because the chatbot it started as wasn't readt for primetime yet while it continued to hone it, they even did a pump and dump SPAC nymag.com/intelligencer/…

...also known as a reverse merger, which was done to fuel further acquisitions, which I presume would have been even more R&D firm acquistion and development of proprietary machine learning. Everything points to a model of using human as a stopgap to an AI content utopia imo

This brings me back to this experiment. You've heard me use the term (which I didn't create) Blavity Blacks. I now say Buzzfeed Blacks because really this style of Black writing is now mainstream. It's bigger than Blavity, Black writers everywhere write like Blavity Blacks now

...in fact Blavity first pitched itself as a Black Buzzfeed for Millenials techcrunch.com/2016/09/12/bla… The advent of Blavity Black style writing is really just the Buzzfeedication of media trend reaching Black journalism

I tried to make the AI yesterday for the thread I did about blue check black writing as contrast write in the style of Ralph Ellison and James Baldwin and it just couldn’t write like great thinkers of the past. It all just came out wrong. But it captures Buzzfeedified wrote easy

So to test the theory further, I used the terms from @tomscocca's anti-Buzzfeed essay, smarm to characterize Buzzfeed and snark to characterize Gawker and asked the AI to write in both styles and here are the results. It literally refused to write the snark.

Once could argue it's because the programmers didn't find such writing constructive but I suspect the other bigger truth is it's harder to write scathing, insightful snarky satire. I think it literally can't even if it wanted to.

Last tweet, another thing that makes me think the goal is to make this the industry standard, soullless rote botlike writing that an AI can easily replicate, is Perretti has openly said he wants to merge all big media majors into a giant monopoly
nytimes.com/2018/11/19/bus…

So looks like this thread is proving prescient. Check today’s news

https://twitter.com/MorePerfectUS/status/1618677568974237697?s=20&t=5tmt6JPOWSokmwgckZlUoQ