Another great piece by @Jake_Browning00 on the limitations of current AI systems.
They are boring https://
t.co/6RS6oeWFgt
RESEARCH
-
Current AI Systems Face Creativity Limitations
By
–
-
New Quantitative Evidence-Driven Approach to AI Safety
By
–
We are working on a a new quantitative & evidence-driven approach to AI safety, beyond what is commonly thought to be possible (and we’re hiring!):
-
New AI Preparedness Team Evaluates AGI Risks Quantitatively
By
–
We are building a new Preparedness team to evaluate, forecast, and protect against the risks of highly-capable AI—from today's models to AGI. Goal: a quantitative, evidence-based methodology, beyond what is accepted as possible:
-
AI Predicts Earthquakes One Week in Advance with 70% Accuracy
By
–
AI could predict earthquakes a week in advance! researchers at UT Austin have developed an AI algorithm that correctly predicted 70% of earthquakes a week before they happened during a seven-month trial in China
-
IBM Granite 13B LLM Training Datasets Revealed
By
–
datasets that IBM used to train its granite.13b LLM: Arxiv, Common Crawl, DeepMind Mathematics, Free Law, GitHub Clean, Hacker News, OpenWeb Text, Project Gutenberg, Pubmed Central, SEC filings, Stack Exchange, USPTO, Webhose, Wikimedia https://
ibm.com/downloads/cas/
X9W4O6BM
… -
End-to-End Spoken Language Model Processing Spectrograms Directly
By
–
Introducing a new approach for spoken language modeling trained end-to-end to directly process spectrograms as both input and output. It can be fine-tuned to generate semantically accurate spoken language for continuation and question answering. → https://
goo.gle/3FzJNvy -
Will larger language models eventually enable critical infrastructure attacks?
By
–
Imagine this prompt: "write me Python code to disable the NYC subway system" obviously gpt-4 can't do this now. It'll refuse, but even if we jailbreak it, it'll answer incorrectly but if we keep training bigger & better language models, won't one eventually be able to do this?
-
Generative AI Pioneers: Exploring Responsibilities and Societal Impact
By
–
Had great fun @StanfordHAI on Tuesday Explored the conference’s theme of New Horizons in Generative AI to an exploration of the Responsibilities of the Pioneer. Text of the talk on my blog (and the video on HAI YouTube soon). https://
blog.shakirm.com/2023/10/respon
sibilities-of-the-pioneer-generative-ai-and-its-sociotechnical-foundations/
… -
Deep learning for brain tumor classification
By
–
Researchers use deep learning to classify brain tumors https://actuia.com/actualite/des-chercheurs-utilisent-le-deep-learning-pour-classifier-les-tumeurs-du-cerveau/
… #AI #artificialintelligence #deeplearning -
Psychological Safety Boosts Team Innovation and Collaboration
By
–
Integrating psychological safety into your team dynamics can help boost innovation and collaboration, research says. Explore strategies to help team members voice opinions, take risks, and be open and honest. Check out these insights and more in @INSEAD https://
rb.gy/h3lxo