Daniel Grady
@danielgrady.net
This is the week of "internet hot takes Claude 4," so …
These examples are all with Claude Sonnet 4 with web search on, extended thinking off.
General theme is that LLMs aren't very good at specific questions about how to use LLMs.
These examples are all with Claude Sonnet 4 with web search on, extended thinking off.
General theme is that LLMs aren't very good at specific questions about how to use LLMs.
Claude
Talk with Claude, an AI assistant from Anthropic
claude.ai
May 29, 2025 at 5:24 PM
This is the week of "internet hot takes Claude 4," so …
These examples are all with Claude Sonnet 4 with web search on, extended thinking off.
General theme is that LLMs aren't very good at specific questions about how to use LLMs.
These examples are all with Claude Sonnet 4 with web search on, extended thinking off.
General theme is that LLMs aren't very good at specific questions about how to use LLMs.
Nice essay from Oliver Reichenstein at iA. Nice essay. People learn by doing and making: although this is an obvious and agreed statement, it should be taken much more literally and much farther than it normally is.
ia.net/topics/maker...
ia.net/topics/maker...
Maker’s Knowledge
We only know what we made ourselves, or: Why we want to know how things are made.
ia.net
May 9, 2025 at 10:31 PM
Nice essay from Oliver Reichenstein at iA. Nice essay. People learn by doing and making: although this is an obvious and agreed statement, it should be taken much more literally and much farther than it normally is.
ia.net/topics/maker...
ia.net/topics/maker...
Oh wow. This morning, I mailed a physical letter to the CA AG requesting action against OpenAI's proposed theft (notforprivategain.org), and today OpenAI announced changes to their plan (www.wired.com/story/openai...). Love seeing this level of efficiency from the USPS!
Not For Private Gain
We write in opposition to OpenAI’s proposed restructuring that would transfer control of the development and deployment of artificial general intelligence (AGI) from a nonprofit charity to a for-profi...
notforprivategain.org
May 5, 2025 at 10:43 PM
Oh wow. This morning, I mailed a physical letter to the CA AG requesting action against OpenAI's proposed theft (notforprivategain.org), and today OpenAI announced changes to their plan (www.wired.com/story/openai...). Love seeing this level of efficiency from the USPS!
Reposted by Daniel Grady
Implement a 100% tariff and ruin a movie
Nineteen Weeks
Nineteen Weeks
May 5, 2025 at 8:44 AM
Implement a 100% tariff and ruin a movie
Nineteen Weeks
Nineteen Weeks
Would be great to get a review from someone who's not just Schmidhubering all over the place. marginalrevolution.com/marginalrevo...
marginalrevolution.com
April 23, 2025 at 5:06 PM
Would be great to get a review from someone who's not just Schmidhubering all over the place. marginalrevolution.com/marginalrevo...
Imagine that AI labs had called it "spew" instead of "Chain of Thought." "When we allow the model to increase its spewage, the final answer tends to be more useful." Does that make Anthropic's new paper more or less interesting? www.anthropic.com/research/rea...
Reasoning models don't always say what they think
Research from Anthropic on the faithfulness of AI models' Chain-of-Thought
www.anthropic.com
April 4, 2025 at 8:59 PM
Imagine that AI labs had called it "spew" instead of "Chain of Thought." "When we allow the model to increase its spewage, the final answer tends to be more useful." Does that make Anthropic's new paper more or less interesting? www.anthropic.com/research/rea...