jcheng

8 ms·

61.
by jcheng 1y ago
Thanks for the thoughtful reply. I agree with all that and I think I was trying to say something similar with my last paragraph. I’ve gotten along with _almost_ every person I’ve worked with, including some pretty challenging personalities.
62.
by jcheng 1y ago
I’ve mostly worked at small startups in my 30 year career, and been fortunate to work mostly with excellent founders and teams. Certainly that experience colors my convictions, as your experience colors yours.
63.
by jcheng 1y ago
It's SO different. It doesn't guarantee that life is good, but politics plays a much smaller role, especially at the very small sizes. If you're objectively awesome, the chances of you being sidelined for political reasons is
64.
by jcheng 1y ago
A huge red flag to me is that the tool calls here are stateless (every tool call is carried out by a new R process) which means the state has to live in the agent’s context, exactly where you don’t want it for so many reasons. For example,
65.
by jcheng 1y ago
Do you have any advice for how to suss out someone's hunger, drive, and scrappiness during the hiring process?
66.
by jcheng 1y ago
If the TTL of one domain’s records are all shorter than the TTLs of another domain’s, what would make downstream and client caching cancel out? Do clients not respect TTLs these days? (In this particular case, I don’t think the TTLs are act
67.
by jcheng 1y ago
The heading says “Generative AI services popularity - Top 10 services based on 1.1.1.1 DNS resolver traffic”
68.
by jcheng 1y ago
Here's my attempt at explaining tool calling: https://youtu.be/owDd1CJ17uQ?si=Z2bldI8IssG7rGON&t=1330 It's an _incredibly_ important concept to understand if you have even a passing interest in LLMs. You need
69.
by jcheng 1y ago
For as long as there have been Jupyter notebooks, there have been people saying "Wait, what??" But there have been way more people using them, and finding them pretty effective despite their inherent lack of soundness. I personall
70.
by jcheng 1y ago
Not at all. The iPod of the day had a click wheel as an input device. The iPhone introduced us to capacitive touch, multitouch, gesture recognition, full web browsing. Huge leap compared to not only the iPod but the PocketPCs of the day. Li
71.
by jcheng 1y ago
Same here, I spent forever on it :(
72.
by jcheng 1y ago
Thank you!
73.
by jcheng 1y ago
Only input, not output. https://modelcontextprotocol.io/specification/2025-03-26/ser...
74.
by jcheng 1y ago
Stuff Made Here https://www.youtube.com/watch?v=WsPHBD5NsS0 Mark Rober https://www.youtube.com/watch?v=Sqr-PdVYhY4
75.
by jcheng 1y ago
And what about using the “back door Roth IRA” to get around the Roth IRA’s income test? What about when the IRS says the back door Roth is allowed? https://www.currentfederaltaxdevelopments.com/blog/2018/7/12.
76.
by jcheng 1y ago
200 examples at most, really?? I have been led to believe that (tens of) thousands is more typical. If you can get excellent results with that few examples, that changes the equation a lot.
77.
by jcheng 1y ago
Or better yet, use CTEs: https://duckdb.org/docs/stable/sql/query_syntax/with.html
78.
by jcheng 2y ago
Those are great, I would watch any one of those movies. Maybe even the "Across the Indiana-Verse" one where they are all pulled into a single dimension.
79.
by jcheng 2y ago
I'm sympathetic in general, but in this case: "You will need git, XCode tools, CMake and libomp. Git, CMake and libomp can be installed via Homebrew" That really doesn't seem like much. Was there more to it than this? Ed
80.
by jcheng 2y ago
It’s pretty astounding to me that this aspect of MCP is not mentioned more. You’re putting a LOT of trust in both the model and the system prompt when you start attaching MCPs that provide unfettered access to your file system, or connect u
81.
by jcheng 2y ago
That makes sense, I was surprised to see the RS3 that high on the list.
82.
by jcheng 2y ago
https://www.caranddriver.com/features/a23319884/lightning-la... The GT-R appears three times, at 75, 76, and 93. The Mustang Dark Horse, Audi RS3, and Supra are all about $60K.
83.
by jcheng 2y ago
Pyodide has numpy, scipy, matplotlib, pandas, and as of last month, even polars. https://github.com/pyodide/pyodide/pull/5282 Pyodide is far from a perfect CPython and even the packages it includes often have
84.
by jcheng 2y ago
Exactly. Was the extra information pushed to the model as part of the query? It’s RAG. Did the model pull the extra information in via a tool call? Agentic search.
85.
by jcheng 2y ago
I agree that retrieval can take many forms besides vector search, but do we really want to call it RAG if the model is directing the search using a tool call? That like an important distinction to me and the name "agentic search"
86.
by jcheng 2y ago
Care to elaborate? I’ve never used it but have heard good things from colleagues who have.
87.
by jcheng 2y ago
Second System Syndrome! :) https://en.m.wikipedia.org/wiki/Second-system_effect
88.
by jcheng 2y ago
I prefer Anthropic's models but ChatGPT (the web interface) is far superior to Claude IMHO. Web search, long-term memory, and chat history sharing are hard to give up.
89.
by jcheng 2y ago
There are several high-level web application frameworks for Python that are based around this concept. It's particularly useful for data-oriented apps that let the user tweak parameters and update (arbitrarily complex) calculations and
90.
by jcheng 2y ago
No, ChatGPT has that feature but not Claude (that I could find) so I pasted my input and its response into a gist. It’s verbatim, I promise.

More ›