Stay #wild and #defiant! #generativeAI and #imageGenerators would never be like you.
Recent searches
Search options
#imagegenerators
0 posts0 participants0 posts today
Nonprofit scrubs illegal content from controversial AI training dataset - Enlarge (credit: Kirillm | iStock / Getty Images Plus)
After S... - https://arstechnica.com/?p=2046573 #childsexualabusematerials #artificialintelligence #stablediffusion1.5 #imagegenerators #aitrainingdata #childsexabuse #aitraining #laion-5b #policy #laion #csam
Ars Technica · Nonprofit scrubs illegal content from controversial AI training datasetAfter backlash, LAION cleans child sex abuse materials from AI training data.
Replied in thread
@db0 @jasonhowell @leo @stablehorde_generator
That's very interesting, and I am experimenting.
Thank you
Replied in thread
do we have a term to distinct reliable #ai tools like #deepblue providing fast & correct responses that transformed how chess masters learn the game, vs nowadays #chatboots #imageGenerators that deliver mostly correct yet mostly including some BS responses?
MastodonTeixi (@teixi@mastodon.social)@emilymbender@dair-community.social
We should communicate to global audiences the distinction between #DeepBlue #AlphaGo and the likes that produced faster while also correct results, thus transformed because being reliable tools for their purposes in chess & go games strategies and results. While the current #LLMs
Is there a term to differentiate #AI #DL #ML that produces reliable results —as for chess gameplay— vs the current #chatbots & #imageGenerators mostly correct but mostly with BS included?
Replied in thread
We should communicate to global audiences the distinction between #DeepBlue #AlphaGo and the likes that produced faster while also correct results, thus transformed because being reliable tools for their purposes in chess & go games strategies and results. While the current #LLMs
Is there a term to differentiate #AI #DL #ML that produces reliable results —as for chess gameplay— vs the current #chatbots & #imageGenerators mostly correct but mostly with BS included?
"#LargeLanguageModels are drafting screenplays and writing code and cracking jokes. #ImageGenerators, such as #Midjourney and #DALLE 2, are winning art prizes and democratizing interior design and producing dangerously convincing fabrications. They feel like magic.
Meanwhile, the world’s most advanced robots are still struggling to open different kinds of doors. As in actual, physical doors. "
#AI Is Running Circles Around #Robotics
https://www.msn.com/en-us/news/technology/ai-is-running-circles-around-robotics/ar-AA19stNO
MSNAI Is Running Circles Around RoboticsMoravec’s paradox strikes again!
Replied in thread
@rasterweb very unlikely. #AI #ImageGenerators are actually useful. For #AI we should spend our time putting up fences and gates more than trying to stop it.
By contrast, #crypto, and #NFT have yet to be useful. They were tech in search of a problem.
Teaching a Robot to Hallucinate - Training robots to execute tasks in the real world requires data — the more, the b... - https://hackaday.com/2023/02/26/teaching-a-robot-to-hallucinate/ #artificialintelligence #machinehallucination #diffusionmodels #imagegenerators #trainingdata #robotshacks #robotics #imagen #ai
HackadayTeaching A Robot To HallucinateTraining robots to execute tasks in the real world requires data — the more, the better. The problem is that creating these datasets takes a lot of time and effort, and methods don’t sc…
*Incredibly prescient cartoon by satirical American newspaper cartoonist HT Webster in 1923 #imagegenerators #AI
“Too easy“—Midjourney tests dramatic new version of its AI image generator | Ars Technica
Ars Technica“Too easy”—Midjourney tests dramatic new version of its AI image generatorVersion 4 offers greater detail and better compositions from simple prompts.