Ian Bicking
LightNews LightNews
ianbicking.org
Ian Bicking
@ianbicking.org
600 followers 850 following 680 posts
Software developer in Minneapolis. Working with applications of LLMs. Previously: Mozilla, Meta, Brilliant.org
Posts Media Videos Starter Packs
ianbicking.org
Ian Bicking @ianbicking.org · 16h
That feels kind of anti-democratic, putting the choice into a smaller, more obscure election…? (Unless it’s just replacing the caucus nomination process.)

I’d rather see a non-partisan RCV primary so the race could be consolidated into two candidates that voters can learn/understand
1
ianbicking.org
Ian Bicking @ianbicking.org · 3d
“Delve” etc seem fairly pointless. They aren’t exceptionally good words, and it feels like an admission of poor diction that they repeat a small set of uniquely cliche words
1
ianbicking.org
Ian Bicking @ianbicking.org · 3d
I don’t really trust there’s a mystery without some declaration. A riddle with no answer isn’t fun
1 1
ianbicking.org
Ian Bicking @ianbicking.org · 4d
Then the next question is: if so, then what?

If something feels like an attack to someone, but you wish it wouldn’t, then what?

(I’d say: don’t waste time wishing it wouldn’t, but there’s more than one answer)
1
ianbicking.org
Ian Bicking @ianbicking.org · 4d
I think I’d enjoy AI art presented in a much more deliberate way:

1. Describe what you are trying to achieve
2. Show the instantiation of the art
3. Reflect on the concrete elements that did or did not succeed

Simply rolling the aesthetic feedback dice gets tiring
1 2
Reposted by Ian Bicking
jzleibo.bsky.social
Joel Z Leibo @jzleibo.bsky.social · 6d
[1/9] Excited to share our new paper "A Pragmatic View of AI Personhood" published today. We feel this topic is timely, and rapidly growing in importance as AI becomes agentic, as AI agents integrate further into the economy, and as more and more users encounter AI.
3 15 52
ianbicking.org
Ian Bicking @ianbicking.org · 4d
I think Democrats had attention and salience around social justice issues, ie the same culture war that MAGA used as the tip of their spear. And Democrats won many of the social justice battles!

But winning those battles doesn’t seem to strengthen the movement. Probably won’t for MAGA either
1
ianbicking.org
Ian Bicking @ianbicking.org · 6d
There's no right answer, but there are _better_ answers. Really it's a discernment process for us to figure out: we don't want the LLM to just "be positive" or "be negative" but instead we have to articulate very carefully what we really do want. (Also we don't know what we want.)
1 1 1
ianbicking.org
Ian Bicking @ianbicking.org · 6d
Is there some universal rubric we should be applying before complimenting the user? Or just hand out compliments at some rate, like "this is in the top 25% of the user's ideas, gold star!" Or concentrate on good ideas and skim over the bad ones?
1
ianbicking.org
Ian Bicking @ianbicking.org · 6d
When I see some clever prompt to make the LLM stop being sycophantic I am reminded of this... the problem isn't really positivity (positivity is actually great!), but a lack of discernment. When it compliments something that doesn't deserve it. But what does "deserve it" even mean?
1
ianbicking.org
Ian Bicking @ianbicking.org · 6d
This comes up with all kinds of LLM discernment. If you think the LLM has a bias in one direction you can tweak the prompting, but you don't get to a "correct" discernment by just making sure the distribution looks correct. If you should accept 50% and reject 50%, it also matters which 50%
1
ianbicking.org
Ian Bicking @ianbicking.org · 6d
Thinking a little more about LLM sycophancy...

In general discernment is very hard to get right. You ask for critique and you'll get critique. You ask for a compliment and you'll get a compliment. There is no "just tell me the truth."
2
ianbicking.org
Ian Bicking @ianbicking.org · 8d
I don't really resent the rate limits, the underlying cost is real. But mostly I'm surprised that at $20/mo (and on the "auto" model setting) Cursor will happily grind for hours every day. And honestly I prefer its results for most coding tasks. (In this case I'm using Claude Code for non-code)
1
ianbicking.org
Ian Bicking @ianbicking.org · 8d
I didn't think I was even using Claude Code that much, and I hit a rate limit. The rate limit also blocks me from using the normal Claude chat interface, which is an interesting choice. OpenAI's Codex similarly conked out fairly early with a rate limit, meanwhile Cursor keeps going and going...
1
ianbicking.org
Ian Bicking @ianbicking.org · 9d
I'm not sure I understand the likely impact of at large seats...? I can imagine a bias towards higher-turnout voting segments. Is it more than that?
1
Reposted by Ian Bicking
python.org
Python Software Foundation @python.org · 9d
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-...
🧵
The official home of the Python Programming Language
www.python.org
130 2.8K 6.3K
ianbicking.org
Ian Bicking @ianbicking.org · 12d
I'm kind of enjoying watching it and trying to reconstruct the surreal tesseract world it embodies.

But I guess what strikes me is how banal the game is. The environment isn't any richer for the AI. Filled with stuff, but just a different kind of background noise.
1 1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
The chatlog, in all its messiness, is a core part of what's happening. It holds the not-yet-actionable information. It holds context, history. It's fully situated, including being situated in a user/assistant relationship. It's unclear what analog exists for a text adventure.
1 1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
Now an LLM can solve that issue pretty well, but what does it parse into? Really there's two things being created by the system: the action and the chatlog. In the context of the chatlog you can parse "change that to eight thirty" into a concrete action.
1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
If you say "set an alarm for 8 o'clock" and there's some {action: "createAlarm", time: "8:00"} structure you can create, then you're good. If you say "change that to eight thirty" and there's no structure to parse that into, then there's no parser that can make that work.
1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
When I was working on voice assistants pre-LLMs I felt the same problem. The parser was only a small part of the problem. You can recognize more phrases, but the only phrases that could matter are the phrases for which some internal target structure existed.
1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
But a better parser isn't that interesting because there's only a couple things a player can do at any moment to advance the game. You can make infinite LLM-backed easter eggs, but that gets old fast.
1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
A lot of people think briefly "wouldn't it be cool to hook an LLM up to a text adventure to parse the input," hoping to solve a real (but minor) problem when the game can't interpret a valid command because it isn't formatted correctly.
1
ianbicking.org
Ian Bicking @ianbicking.org · 14d
It appears to be freedom because you can't see the map of possible actions. Anything you type _could_ work. This makes the game feel open but it's an illusion, your mind imagining freedoms you have not yet pursued; but also you cannot pursue them.
1 2
ianbicking.org
Ian Bicking @ianbicking.org · 14d
In practice I don't think text adventures offer much freedom. You can't dream up new ways to solve puzzles. Picking up and dropping inventory items is not a deep system. Determining the correct verb to use can be a puzzle of its own, but it's not freedom.
1