À San Francisco, les taxis autonomes sans chauffeurs pilotés par Intelligence Artificielle sont autorisés à circuler 24h sur 24h L’ensemble de la Californie devrait suivre Souhaitez-vous que la France fasse la même chose ? Ou Voulez-vous sauver les @TAXISG7 ?
AI
-
Texture Datasets Updated for Generative AI Training
By
–
I updated my textures datasets, both Natural Ground pack and Urban Surfaces pack. There are now 17 and 16 sets in each one respectively for a total of 432 images at 4K cc-by-sa-nc. Fun for training domain-specific generators! https://
huggingface.co/datasets/textu
redesign/td01_natural-ground-textures
… https://
huggingface.co/datasets/textu
redesign/td02_urban-surface-textures
… -
Voice AI Ethics: Latent Space Similarity and Publicity Rights
By
–
It's partly a question of dataset, but also a question of how similar the voice "code" is in the latent space. Maybe they picked that point to be similar intentionally, or it was an accident… either way, if it's similar moral rights and Right Of Publicity are infringed.
-
AI for the Developing World: Harnessing Power
By
–
Artificial Intelligence for the Developing World. How to Harness the Power of AI. #ChatGPT #OpenAI #AI #ArtificialIntelligence #ChatGPT4 #MachineLearning #DeepLearning #Tech #Innovation #FutureTech #ML #IoT #technews #techforgood https://
foreignaffairs.com/world/artifici
al-intelligence-poor
… -
ChatGPT Expands Custom Instructions Feature to Free Users
By
–
ChatGPT expands ‘Custom Instructions’ to free users. https://
artificialintelligence-news.com/2023/08/10/cha
tgpt-expands-custom-instructions-free-users/
… #ChatGPT #OpenAI #AI #NLP #ArtificialIntelligence #ChatGPT4 #MachineLearning #DeepLearning #Tech #Innovation #FutureTech #ML #IoT #Programming -
Context Extraction is Key to Accurate LLM Answers
By
–
answering is not a problem. llama2-70b works quite well. extracting the right context is the bigger issue 🙂 once the context is correct, answers are automatically correct even with smaller LLMs 🙂
-
Semantic search fails at scale beyond toy examples and familiar data
By
–
ive been trying to make semantic search work on 150k documents for a while. what ive found so far is that it doesnt work. it only works for toy examples, very small datasets and data which the LLM has more or less already seen and that's why it answers well from it.
-
Future of Jobs in AI, Sustainability and Deglobalization
By
–
The future of jobs in the age of #AI, #sustainability and deglobalization https://
bit.ly/3WaL7g6 via @wef #FutureofWork -
Media Coverage of Effective Altruism and AI Influencers Criticized
By
–
And @TIME
. Who can forget their cove r story doing free PR for effective altruists and soliciting op-Ed’s from other cult leaders like Yudkowsky? Oh and let’s not forget @TEDTalks
, where is was also platformed It’s the same cycle over and over. -
NYT platforms AI philosophers amid ethics concerns
By
–
Not to mention @nytimes inexplicably platforming random philosophers like Bostrom and MacAskill. Do no eugenicists get that type of platform over there? They even did a whole interview with Bostrom a few months after his email and worse non apology came out.