fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett) header
fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett) profile picture

๐Ÿฆธ๐Ÿปโ€โ™‚๏ธ @superhuman ร— @gptzeroai by day ๐Ÿก @abodefyi / https://xitter.gq/t.co/IkC1PP58XB by night โœ๏ธ words https://xitter.gq/t.co/XojapbmiJ8 โฌ…๏ธ prev founding eng @lexdotpage

london ๐Ÿด๓ ง๓ ข๓ ฅ๓ ฎ๓ ง๓ ฟ
Joined February 2009
Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)Photo by fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ (@fredrivett)
well thatโ€™s unexpected
20
3
5
435
92,797
the odds neuralink is the mark of the beast are disconcertingly high
1
156
fred rivett ๐Ÿ‡ฌ๐Ÿ‡ง๐Ÿ“ˆ retweeted
person who thinks AI will prove the Riemann hypothesis in 2026 but that Apple wonโ€™t deliver an LLM-powered Siri until 2027
182
517
191
6,297
525,981
for those who think ai detectors are regularly mischaraterising human work as ai, theres some free money over here
Today, we are announcing GPTZero's challenge for AI detection with USD $13,000 in rewards - $1000 for the first 750-word false positive for pre-2020 writing -$12,000 in research awards and grants -$100 for any false positive over 300 words, as long as there is evidence of human writing or paraphrasing Why are we doing this? - We want to raise the bar on benchmarking AI detectors. Our latest model outperforms Pangram 4 in 9 out of 11 major benchmarks. we want to put that to the test in the real world - we want to keep our team accountable, both to pay out and learn from any false positives - most importantly, we want to actually fund interesting research on how AI is changing the internet. We are also giving researchers free access to our API and MCP credits, both upfront and with 100% refunds on any cost for their research. - researchers and writers are spending hundreds of dollars testing AI detection models, we want this cost to come from us, and not their pockets. Full rules of the challenge, and steps to access API and research credits are listed on gptzero[.]me/bounty
3
121
pretty wild we used to hand type every line of code
2
154
the depths of craft going on over at @dessn_ai are insane
built a custom gaussian splat renderer just to put this flower on our waitlist page the old one took too long to load. went from 1.3 MB of rendering code to 9 KB, and it starts playing twice as fast all this for a page with one button
1
1
4
313
if any design studios are looking for a new brand designer, i know a top-tier mid-senior designer who's open to new opportunities, london or remote
1
185
self-deprecating fun is the vibe
๐Ÿšจ MUSE FEATURE ALERT ๐Ÿšจ you can now try on clothes in muse before you buy them ๐Ÿ‘• ๐Ÿ‘— ๐Ÿ‘– muse will offer to show you how you look in something while youโ€™re shopping for it. trying things on in person? as if....
67
wild i'm roughly 2-4x more productive than this time last week
1
1
267
if your workflow looks anywhere close to what it was 6mo ago, skill issue
77
how are y'all organising your personal todo's these days?
71
is nothing sacred anymore?
JUST IN: PlayStation network down for many users as GTA 6 set to release worldwide in 10 hours.
2
84