BLUE
Profile banner
DO
Disappointed Optimist
@disappoptimism.bsky.social
I’m that guy that looks just like this that did that tweet that made you laugh that one time. UK based.
2.5k followers409 following4k posts
DOdisappoptimism.bsky.social

On the other place saw this tweet and it made me sad. The age of the internet peeps is over; it’s the machines turn now

21

Hhughster.bsky.social

The era of the 90s wild west "open internet" might be over. The era of the whitelisted, trusted-source-only "closed internet" might be just beginning.

0
Rrucolaspacecat.bsky.social

📌

0
Ccdpositive.bsky.social

Spoken language is no longer a human activity. Maybe it hasn't been for a while (hey, Siri?). 🥹

0
CMmarlowechris.bsky.social

This is tragic

0
CPcapetrov.bsky.social

With alt text

Tweet by Daniel Feldman @d_feldman

The widely-used wordfreq database of English word frequencies will no longer be updated.

Screenshot of an article:
Generative Al has polluted the data

I don't think anyone has reliable information about post-2021 language usage by humans.

The open Web (via OSCAR) was one of wordfreq's data sources. Now the Web at large is full of slop generated by large language models, written by no one to communicate nothing. Including this slop in the data skews the word frequencies.

Sure, there was spam in the wordfreq data sources, but it was manageable and often identifiable. Large language models generate text that masquerades as real language with intention behind it, even though there is none, and their output crops up everywhere.

19/09/2024
0
outeast.bsky.social

I hadn't even thought of this impact, though it's self-evident. How depressing. LLMs really have borked the web in so many ways.

0
DOdisappoptimism.bsky.social

Humanity managed to dirty the online ecology so badly that Google AI is giving out terrible answers

4
Ppenbird42.bsky.social

"Can't you train an LLM to filter out the LLM garbage?" Oh shut up

1
KFkevfquinn.bsky.social

Link to the source (and a reminder linking to the source does *not* reduce reach on Bluesky, please find and add the source link when posting screenshots!) github.com/rspeer/wordf...

0
EWeddwilson.bsky.social

When the AI is trained on material increasingly generated by AIs. . .When the AI is trained on material increasingly generated by AIs. . .

2
Profile banner
DO
Disappointed Optimist
@disappoptimism.bsky.social
I’m that guy that looks just like this that did that tweet that made you laugh that one time. UK based.
2.5k followers409 following4k posts