robwwilliams

9 ms·

91.
by robwwilliams 1y ago
Hilarious, and so close to Claude default mode (well yes, parody lol thereof). Try this pre-prompt: Please respond in mode of Ernest Hemingway “You’re right. When someone explains why they’re explaining something, it goes in circles. Like a
92.
by robwwilliams 1y ago
Not sure I understand your point. Add a qualifier if you want. It is not taking money as much as responding to a request for proposals.
93.
by robwwilliams 1y ago
Worth drawing a distinction between governmental support for science and for the humanities. The first does a lot of relative low mark-up contract work requested by governmental agencies. Governments and all of us would like to see cancer a
94.
by robwwilliams 1y ago
Small sample size of about 1500 Amish men divided across 4 cohorts, all exposed to the great depression. Entry age minimum of 25 years. Sorry, but this is really marginal science. There are much stronger demographic and statistical studies
95.
by robwwilliams 1y ago
HN is a good model. There is philosophically no credible way to enforce balance—-no common reference points of truth with capital T. But accepting a plurality of views that are shared dispassionately may be achieved. Hacker News usually suc
96.
by robwwilliams 1y ago
All I can say is that Claude Sonnet has been an amazing help for me in reviewing published papers and in “pre-reviewing” my own grant applications. And I trust it to be much more impartial than my polarized peers.
97.
by robwwilliams 1y ago
This! Today I asked Claude Sonnet to read the Wikipedia article on “inference” and answer a few of my questions. Sonnet responded: “Sorry, I have no access.” Then I asked it why and it was flummoxed and confused. I asked why Anthropic did n
98.
by robwwilliams 1y ago
The jump to 1 million token length context for Sonnet 4 plus access to internet has been a game-changer for me. And somebody should remind Anthropic leadership to at least mirror Wikipedia; better yet support Wikipedia actively. All of the
99.
by robwwilliams 1y ago
You are right given the context of the Science target article, which provides an example of proactive uses if AI to suggest innovative work.
100.
by robwwilliams 1y ago
This is disallowed by NIH for confidentiality concerns. But reviewer are free to submit the applicants paper to an LLM for commentary. Useful to detect traces of p-hacking that require careful teading.
101.
by robwwilliams 1y ago
Massively burdensome? No, it is an absolutely essential process that over the last 80 years of “industrial academic research” has brought us wonders and insights. Three to four months to design. write, and refine a 13 page document is work
102.
by robwwilliams 1y ago
Not my experience with AI at all. In what way do you mean AI makes binary decisions? I often ask AI systems for pros and cons and thy do as good a job as I would in many situations in which I am knowledgeable.
103.
by robwwilliams 1y ago
This is perhaps a good argument for AI-enhanced reviews. in my experience since 1982, reviewers will pick you to death. Peers know the real complexity. Sure, you will occasionally get reviewers who completely miss the point, but it is usual
104.
by robwwilliams 1y ago
Why do you say that writing grant applications is a waste of time? For me it is a time to dive deeply into the state-of-the-art and figure out better ways to address important questions. I do not see writing grants as a trivial core, but as
105.
by robwwilliams 1y ago
Too pessimistic to the point of being apocalyptic. AI can be used as an editorial assistant, research assistant, and copy editor. This is a huge benefit already. AI systems are also almost brilliant in helping to tighten text. Try compressi
106.
by robwwilliams 1y ago
Odd that you are being down-voted for pointing out the easy opt-out option. I need this opt-in feature. I suppose the polarity of action could be a factor.
107.
by robwwilliams 1y ago
Is there a reason you cannot opt out? Or don’t you trust Anthropic’s opt-out implementation.
108.
by robwwilliams 1y ago
Yes, I share your take on this change—a way to improvements in many domains. Agree that the fact that these improvements will accrue within in a proprietary for profit. But still a net positive fir my work. Give me a FOSS LLM with Claude 4
109.
by robwwilliams 1y ago
You are right. To the best of my knowledge there has notbeen a way to share conversations with Anthropic, and therefore with future versions of Claudes. There was no Opt-in option as far as I know. Straighten me out if I am wrong. I need t
110.
by robwwilliams 1y ago
Ditto: was delighted to see this as an option. Am I missing some details? I do not understand the comments in “rug pulls”.
111.
by robwwilliams 1y ago
You want them to flush your conversations on your own schedule. That I can understand. If you delete a conversation it should be DELETED.
112.
by robwwilliams 1y ago
It is optional. I want the sharing option.
113.
by robwwilliams 1y ago
Agreed. I actually want and need Claude and other LLMs to learn from my interactions with them. Exceedingly frustrating that LLMs do not yey hold a longterm memory or embedded trace of conversations per user. I have been asking Anthropic fo
114.
by robwwilliams 1y ago
I find Maturana’s definition satisfactory. Best overview in his book with Francesco Valera: Autopoiesis and Cognition: The Realization of the Living (1980) https://en.wikipedia.org/wiki/Autopoiesis_and_Cognition:_The...
115.
by robwwilliams 1y ago
Fascinating! The article should have at least tipped its hat to mitochondria: >But unlike a virus, Sukunaarchaeum has its own ribosomes, cellular structures that synthesize proteins, and it can replicate itself without the help of a host
116.
by robwwilliams 1y ago
Sorry, yes “gynasium” is the academical track. There are two “lower” tracks. All decided when you are 11 years old! Not sure uf this is still SOP but was in the 60s through ???.
117.
by robwwilliams 1y ago
Wow, a baseline environment everyone in America shares. Come visit Memphis.
118.
by robwwilliams 1y ago
Grab Claudet and Claudette quick!
119.
by robwwilliams 1y ago
Wait, aren’t you describing Dropbox?
120.
by robwwilliams 1y ago
Great overview of one key problem in prediction from genotypes. So many more to go: oligogenic and polygenic models and then epistasis and gene-by-environmental interactions, epigenetic controls, translation controls …

More ›