Topics
late
AI
Amazon
Image Credits:Bryce Durbin / TechCrunch
Apps
Biotech & Health
clime
Image Credits:Bryce Durbin / TechCrunch
Cloud Computing
Commerce
Crypto
An image reportedly generated by red_panda.Image Credits:Deedy Das(opens in a new window)
Enterprise
EVs
Fintech
Image Credits:Artificial Analysis
fund raise
Gadgets
punt
Another image reportedly from red_panda.Image Credits:Neuralithic(opens in a new window)
Government & Policy
ironware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
societal
blank
Startups
TikTok
Transportation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
video
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
Update : The framework ’s been revealed asRecraft ’s Recraft v3 . The original story follow :
A cryptic young simulacrum generation good example is beat models from Midjourney , Black Forest Labs , and OpenAI on the crowdsourced Artificial Analysis benchmark .
The model , which goes by the name “ bear cat , ” is around 40 Elo points ahead of the next - well - ranking fashion model , Black Forest Labs’Flux1.1 Pro , on Artificial Analysis ’ text - to - prototype leaderboard . Artificial Analysis uses Elo , a rank system originally developed to calculate the relative skill grade of cheat players , to liken the performance of the various model it tests .
Similar to the community AI benchmarkChatbot Arena , Artificial Analysis ranks model through crowdsourcing . For image models , Artificial Analysis take two mannikin at random and feed them a unequaled prompting . Then it presents the prompt and resulting images , and users choose which they think better reflects the prompt .
Granted , there ’s some bias in this balloting process . Artificial Analysis ’ voter are AI partizan , for the most part , and their choices might not reflect the preferences of the wide community of generative AI users .
But panda is also one of the substantially - perform models on the leaderboard in term of its coevals amphetamine . The poser takes a median of around 7 seconds to generate an range of a function — more or less two times faster than OpenAI’sDALL - E 3 .
So , where did red_panda make out from ? Which company made it ? And when can we expect it be release ? All good questions . AI labs increasingly use community bench mark todrum up anticipationahead of an announcement , though , so it might not be long before we find out .
