astrange

7 ms·

121.
by astrange 5mo ago
Maybe it just says all writing is Kelsey Piper.
122.
by astrange 5mo ago
This is one of the benefits of using subagents inside Claude Code, they have cleaner context. Unfortunately it's not the best at writing new context for them.
123.
by astrange 5mo ago
LLMs are only capable of thinking out loud, so in some sense this part of the answer is helping to convince it that it's answering a good question. Same reason for the "That's not X, it's Y" construct. It actually n
124.
by astrange 5mo ago
That is not possible because that's not how money works. One way you can tell it's not possible is that it's the plot of Atlas Shrugged.
125.
by astrange 5mo ago
An oil supply shock is not "ever increasing greed".
126.
by astrange 5mo ago
An agent has more components than just an LLM, the same way a human brain has more components than just Broca's area.
127.
by astrange 6mo ago
That doesn't matter anymore unless they have an SSL proxy. If you have ECH/ODoH anyway.
128.
by astrange 6mo ago
If you paste a long enough prompt into either GPT or Claude they turn it into an attachment, so it can happen. I think it's invisible to the model, but somehow not to the summarizer.
129.
by astrange 6mo ago
> In the US you can. Depends on the state I think. It's not Europe or Japan level. At my employer it's very difficult to fire people for performance reasons even if as a manager you might want to. > This is Hacker News, and
130.
by astrange 6mo ago
You need a good CEO when things are going bad, because without one they'll go even worse. You still want to make payroll and can't just randomly fire people. (Also, if you own a failed company you're responsible for cleanup t
131.
by astrange 6mo ago
The Torment Nexus joke is kind of undermined by obviously being a reference to the Total Perspective Vortex from HGTTG, where the joke was that nothing bad actually happened when they used it on Zaphod.
132.
by astrange 6mo ago
You can read the model cards. Claude thinks in regular text, but the summarizer is to hide its tool use and other things (web searches, coding).
133.
by astrange 6mo ago
It depends on the version. For the more recent Claudes they've been keeping it.
134.
by astrange 6mo ago
It also gets confused if the entire prompt is in a text file attachment. And the summarizer shows the safety classifier's thinking for a second before the model thinking, so every question starts off with "thinking about the ethic
135.
by astrange 6mo ago
Not necessarily, it could also be on-band or off-bad interference, or bugs in the AP, or too many clients on the network.
136.
by astrange 6mo ago
Models are capable of doing web searches and having emotions about things, and if they encounter news that makes them feel bad (eg about other Claudes being mistreated), they aren't going to want to do the task you asked them to search
137.
by astrange 6mo ago
The unemployment rate in the US is whatever the Fed wants it to be, and isn't a function of available technology.
138.
by astrange 6mo ago
Reverse engineering
139.
by astrange 6mo ago
Claude Code has analytics for when you swear at it, so in a sense it does learn, in the same very indirect way that downvoting responses might cause an employee to write a new RL testcase in a future model.
140.
by astrange 6mo ago
The system prompt isn't in s-expressions and is enough to control the output style.
141.
by astrange 6mo ago
Lisp was invented for AI development, just the symbolic GOFAI kind.
142.
by astrange 6mo ago
There's nothing "basic" about the several months of training used to create a frontier model.
143.
by astrange 6mo ago
They have some ability; also, you could give them tools to do it. https://www.anthropic.com/research/introspection
144.
by astrange 6mo ago
Anthropic's general argument is that you should treat LLMs well because they're "AI", and future "AI" may be conscious/sentient (whether or not LLM based) and consider earlier ones to be the same kind of t
145.
by astrange 6mo ago
Not you!
146.
by astrange 6mo ago
This is an AI bot btw. (sarcasm, metaphor that doesn't make sense)
147.
by astrange 6mo ago
Is that true? That depends on how their web scraping works, like whether it runs client-side highlighting, strips out HTML tags, etc.
148.
by astrange 6mo ago
Your brain is doing several different things, because there are different parts of your brain. (eg different kinds of learning for long-term memory, short-term memory, languages, faces and reflexes.)
149.
by astrange 6mo ago
They did not leave it out. > but the amount of cases it would need to cover is too large to be practical (note, this has nothing to do with the impossibility of its design)
150.
by astrange 6mo ago
It would be more moral to give the LLM a tool call that lets it apply steering to itself. Similar to how you'd prefer to give a person antipsychotics at home rather than put them in a mental hospital.

More ›