jsw97

6 ms·

61.
by jsw97 7mo ago
Definitely, you have to weigh the benefit of experience against the cost of revelation. (And all the other costs of course.)
62.
by jsw97 7mo ago
The author explains that this problem is actually adversarial, in the sense that the attacker gets to observe defenses and allocate warheads and decoys accordingly. Thinking of our current circumstances, this suggests another cost of war: o
63.
by jsw97 7mo ago
All due respect this is by no means one of the most influential papers in economics.
64.
by jsw97 7mo ago
Or, less securely, hash the tests and check the hash with a hook, post tool use. Or a commit hook.
65.
by jsw97 7mo ago
This seems like the most intelligent response, in some ways, because it’s funny and tweaks the user. I could see myself replying like this.
66.
by jsw97 8mo ago
In the glorious future, there will be so much slop that it will be difficult to distinguish fact from fiction, and kompromat will lose its bite.
67.
by jsw97 8mo ago
I ran into this yesterday and disabled it by changing permissions on the project’s memory directory. Claude was unable to advise me on how to disable. You could probably write a global hook for this. Gross though.
68.
by jsw97 9mo ago
This is scary and it affects the degree to which I invest in building Claude-specific tooling, either code or in my brain. You can never guarantee that a dangerously-skip-permissions session is going to stay on the rails, what flags it migh
69.
by jsw97 1y ago
I believed the original story. Now I don’t. So it helped me.
70.
by jsw97 1y ago
You could do that -- my understanding is that MCPs give Claude less to "think about" than having to use another program correctly, and therefore avoid context clutter. I could be wrong.
71.
by jsw97 1y ago
I use the Gemini mcp and my CLAUDE.md has instructions to consult Gemini for strategy. Seems to work well. I have the lowest Claude plan so I don’t know how this would work vs Opus, for example. Separately, I have been meaning to implement
72.
by jsw97 1y ago
He switched out his laptop.
73.
by jsw97 1y ago
I know this was a throwaway, but finding prior art for a large group of existing patents would be a cool application.
74.
by jsw97 3y ago
It is not. That approach was tried and rejected, hence the problem.
75.
by jsw97 3y ago
If this were true you would expect an announcement effect for disclosures. Testable.
76.
by jsw97 4y ago
Don't pay fees.
77.
by jsw97 4y ago
Russia has committed so many own goals in this war that the fact that it is against Russia's interests to blow up Nord Stream actually increases the probability that they did it.
78.
by jsw97 4y ago
They already know about us.
79.
by jsw97 5y ago
Seemingly narrow misses often cause large declines. The reason is that management has a lot of ways to guide analysts to reasonable numbers. Like it or not, companies also have ways of stretching to make it work. If the company can’t mee
80.
by jsw97 5y ago
Standards of proof are different.
81.
by jsw97 5y ago
By the time lead is detectable it has done meaningful damage. Yes it’s better to prevent more at that point so don’t stop testing. But testing is a poor prevention mechanism. Lead also does damage below the threshold of detection.
82.
by jsw97 5y ago
I wonder if the Boards of Canada vibe is a conscious choice.
83.
by jsw97 5y ago
Paul Samuelson wrote an article in 1979 on this. It contains only one-syllable words, except for the last word, which is "syllable". Title: "Why We Should Not Make Mean Log of Wealth Big though Years to Act Are Long."
84.
by jsw97 5y ago
"We really trust our rats, because very often after clearing a minefield, our teams will play a game of soccer on the cleared field to assure the quality of our work," he said.
85.
by jsw97 6y ago
As it happens, this has been done before: https://skrekkogle.com/projects/postpost/ Not that this incarnation isn't clever as well.
86.
by jsw97 6y ago
Amen to this. The effortless translation of matrix expressions into code is a huge positive for Matlab/Octave and negative for Python. Have not worked with Julia, and I wonder what the experience is there.
87.
by jsw97 6y ago
Event studies have been around since the late 1960s.
88.
by jsw97 7y ago
I can add to this praise. I have been using key-chord-mode in emacs for the last 7 years. Love it.
89.
by jsw97 8y ago
Yes, that's actually the causality they suggest in the paper.
90.
by jsw97 8y ago
This is apparently a common practice, although rarely this blatant. Paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2479542 Paper finds that "firms that 'cast' their conference calls by dispr

More ›