Thank you @preskill I do think @caltech is a very special place and a perfect environment to pursue interdisciplinary research in #AI + Science
AI
-
OpenAI Mission: Ensuring AGI Benefits Humanity
By
–
Mission of OpenAI is to ensure AGI benefits all of humanity. Key ingredients we think will be important:
-
Three Essential Requirements for Safe AGI Future Development
By
–
things we need for a good AGI future: 1) the technical ability to align a superintelligence
2) sufficient coordination among most of the leading AGI efforts
3) an effective global regulatory framework including democratic governance -
Bengio and Russell: Knowledge vs Building AI Systems
By
–
Bengio and Russel surely know AI. But I don’t think they are building AI systems. And I don’t mean this in any negative way.
-
Breakthroughs Don’t Always Equal Commercial Success
By
–
A breakthrough can be large or small. A small breakthrough can be legit but still not get a drug over the finish line. Breakthroughs and commercial success are not the same thing.
-
ControlNet Checkpoint Conversion and Diffusers Library
By
–
for the .ckpt for the control network you can use the ones in this repo: https://
huggingface.co/thibaud/contro
lnet-sd21
… just realized the diffusers library has scripts to convert from ckpt to diffusers but not in reverse. -
Batch File Ingestion into Delta Lake Using COPY INTO
By
–
Q: How can I perform batch file ingestion into #DeltaLake from cloud object stores? A: COPY INTO! COPY INTO is a #SQL command that ingests files with exactly-once semantics when executed multiple times. Learn more
-
GPT’s inability to unify concepts across languages enables jailbreaks
By
–
imo this jailbreak highlights a unique lack of understanding of "unified concepts" by GPT if GPT analogously mapped concepts to entities regardless of language, it would be able to shut down my Greek adversarial prompt like it did when I asked the same prompt in English
-
World Dictatorship for AI Control: Risks and Concerns
By
–
Some people even suggested a centralized world dictatorship so that AI development can be coordinated and controlled. I don't know what can go wrong in this scenario (?)
