Topics
Latest
AI
Amazon
Image Credits:Richard Bord/WireImage / Getty Images
Apps
Biotech & Health
Climate
Image Credits:Richard Bord/WireImage / Getty Images
Cloud Computing
commercialism
Crypto
Enterprise
EVs
Fintech
Fundraising
Gadgets
bet on
Government & Policy
computer hardware
Layoffs
Media & Entertainment
Meta
Microsoft
concealment
Robotics
security measures
societal
outer space
inauguration
TikTok
deportation
speculation
More from TechCrunch
consequence
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
Elon Musk ’s Grok unloose anew AI image - genesis featureon Tuesday night that , just like the AI chatbot , has very few safeguards . That mean you may generate imitation images of Donald Trump smoking marijuana on the Joe Rogan show , for example , and upload it straight to the hug drug platform . But it ’s not really Elon Musk ’s AI troupe powering the madness ; rather , a new startup — Black Forest Labs — is the outfit behind the controversial characteristic .
The collaboration between the two was give away when xAIannouncedit is turn with Black Forest Labs to power Grok ’s image generator using its FLUX.1 modelling . An AI image and picture startup that found on August 1 , Black Forest Labs appears to commiserate with Musk ’s vision for Grok as an “ anti - woke chatbot , ” without the stern guardrails found in OpenAI ’s Dall - E or Google ’s Imagen . The social metier site is alreadyflooded with exorbitant imagesfrom the fresh feature .
It looks like Grok ’s AI image author is here . And yes , as expected , very few safeguardspic.twitter.com/AB3IO7A3Dc
Black Forest Labs is found in Germany and recently come out of stealth with $ 31 million in ejaculate funding , led by Andreessen Horowitz , according to apress release . Other far-famed investor include Y Combinator CEO Garry Tan and former Oculus CEO Brendan Iribe . The inauguration ’s Centennial State - founders , Robin Rombach , Patrick Esser , and Andreas Blattmann , were formerly research worker who help create Stability AI ’s Stable Diffusion models .
According toArtificial Analysis , Black Forest Lab ’s FLUX.1 models surpass Midjourney ’s and OpenAI ’s AI image generators in terms of quality , at least as ranked by user in theirimage orbit .
The inauguration says it is “ hold our models available to a wide-cut audience , ” with open root AI persona - generation mannequin onHugging FaceandGitHub . The company says it project to create a text - to - video exemplar soon , as well .
Black Forest Labs did not immediately answer to TechCrunch ’s request for comment .
Oh my god . Grok has utterly no filters for its image generation . This is one of the most foolhardy and irresponsible AI effectuation I ’ve ever seen.pic.twitter.com/oiyRhW5jpF
In its launch release , the fellowship says it direct to “ enhance trust in the safety of these models ” ; however , some might say the flood of its AI generated image on X Wednesday did the opposite . Many images exploiter were able to make using Grok and Black Forest Labs ’ tool , such as Pikachu hold an assault rifle , were not able to be re - created with Google or OpenAI ’s picture author . There ’s certainly no doubt that copyrighted imagery was used for the poser ’s training .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
That’s kind of the point
This lack of safeguards is likely a major reasonableness Musk chose this confederate . Musk has made clear that he conceive guard actually make AI models less good . “ The danger of educate AI to be waken — in other word , lie in — is deadly , ” said Musk in atweetfrom 2022 .
Board director of Black Forest Labs , Anjney Midha , posted on Xa series of comparisonsbetween icon generated on daylight one of launch by Google Gemini and Grok ’s Flux quislingism . The yarn highlights Google Gemini ’s well - documented subject with creating historically exact images of people , specifically by come in racial multifariousness into images unsuitably .
“ I ’m glad@ibaband squad took this earnestly and made the right choice , ” said Midha in atweet , referring to FLUX.1 ’s seeming shunning of this offspring ( and mentioning the account of xAI precede researcher Igor Babuschkin ) .
Because of this flub , Google apologized and wrench offGemini ’s ability to generate imagesof people in February . As of today , the company still does n’t get Gemini generate images of people .
A firehose of misinformation
This general lack of safeguards could cause problems for Musk . The X program draw criticism when AI - generated deepfake explicit figure of speech exemplify Taylor Swiftwent viral on the platform . Besides that incident , Grok generateshallucinated headlinesthat appear to drug user on X almost weekly .
Just last hebdomad , five secretaries of stateurged X to stop spreading misinformationabout Kamala Harris on X. Earlier this month , Muskreshareda video that used AI to clone Harris ’ interpreter , making it look as if the frailty chairperson admit to being a “ diverseness hire . ”
Musk seems wrapped on countenance misinformation like this pervade the chopine . By allowing substance abuser to post Grok ’s AI simulacrum , which seem to miss any watermark , directly on the platform , he ’s fundamentally open up a firehose of misinformation pointed at everyone ’s X newsfeed .
