Fuwn (@Fuwned) header
Fuwn (@Fuwned) profile picture

Fuwn

@Fuwned

AI/ML Engineer, Agent Wrangler; forcing gradient descent to larp intelligence. Open to roles!

 Ground Zero
github.com/Fuwn
Joined August 2017
  • Tweets151
  • Following501
  • Followers968
50Media
Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)Photo by Fuwn (@Fuwned)
Pinned Tweet
Fuwn (@Fuwned) profile picture
Fuwn
@FuwnedNov 1
Back in early 2026, I tried to force memory into all my agents: persistent memory across ChatGPT, Codex, Claude, Claude Code, etc. Now I value ephemeralism far more.
7
3
133
Happy Halloween! github.com/Fuwn
Image from Fuwn (@Fuwned)
48
Fuwn retweeted
Tibo (@thsottiaux) profile picture
Tibo
@thsottiauxOct 29
Day 25/ We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today. We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage. And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.
7,386
2,679
3,628
38,722
7,046,021
We're back and Botting.
Fuwn (@Fuwned) profile pictureFuwn
@FuwnedOct 29
Seems like there's a small Grok Bot incident, but only for certain Bots? One of my Bots keeps throwing the error "We can't confirm your billing for this request", while all my other Bots work fine. The Grok website shows no error, and X Chat shows "You’re out of Bot usage for now."
1
39
Seems like there's a small Grok Bot incident, but only for certain Bots? One of my Bots keeps throwing the error "We can't confirm your billing for this request", while all my other Bots work fine. The Grok website shows no error, and X Chat shows "You’re out of Bot usage for now."
1
1
2
223
SpaceXSI Status (status.x.ai/) still shows everything is fine, like 20 minutes after I was initially hit.
SpaceXSI Status

SpaceXSI Status

Current status of SpaceXSI products and APIs.

status.x.ai
18
I will never stop being impressed by how terrible models are at projecting ETAs. Opus 5.5 genuinely thought it would take “3600-4000 agent hours”. It took 16.
49
Fuwn retweeted
lyra (@_lyraaaa_) profile picture
lyra
@_lyraaaa_Oct 24
if you have a qwen3.5 4b around, please try steering neurons L14 N8949 +1 and L13 N7956 -1 now you have evil qwen recipe: dose × median(‖resid‖ leaving layer L, tokens 1:) / ‖W_down[:, n]‖` at the down_proj at every input
514
1,325
777
22,394
4,517,331
Fuwn retweeted
Wyper (@WyperXx) profile picture
Wyper@WyperXxOct 25
61
2,736
66
27,777
448,472
As a Codex app enthusiast, I finally got around to diving deep into T3 Code. Just use T3 Code.
1
3
215
Pumpkin Bot has been achieved internally (restart your app)
167
119
52
2,362
92,167
Fuwn retweeted
Tibo (@thsottiaux) profile picture
Tibo
@thsottiauxOct 21
Day 16/ You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation or resuming after a few hours and are now guaranteed to hit the cache.
OpenAI Developers (@OpenAIDevs) profile pictureOpenAI Developers
@OpenAIDevsOct 20
You can now choose 12-hour prompt caching in the API. Persistent agents, like dots, reuse context throughout the day. You can keep shared instructions, tools, and context cached longer and pay the lower cached-input rate, even when requests are hours apart.
2,607
682
1,092
15,468
3,152,500
oh now it’s we 😭
Image from Captain Insight (@CaptainInsightX)
49
184
9
11,551
108,688
Benchmarked Muse Wonder against GPT-6.1 Sol on a Forth interpreter. Muse Wonder is very fast and a little sloppy. GPT-6.1 Sol is slow and careful. For anything I have to review, I’ll take slow.
157
Here are a few of my burn rates from using the GPT-6 family of models on long-horizon tasks. GPT-6.1 Sol (High, Fast): ~0.26%/h GPT-6 Sol (XHigh): ~0.89%/h GPT-6 Sol (Max): ~0.96%/h GPT-6 Astra (Max): ~1.94/h GPT-6 Astra (Ultra): ~9.5%/h GPT-6 Sol (Ultra): ~18-19%/h
136
I'm loving GPT-6.1 Sol. It's been as tenacious as GPT-6 Astra in my work, and it's dirt cheap.
80
Fuwn retweeted
Railway (@Railway) profile picture
Railway
@RailwaySep 28
hi! i'm raila-chan, an AI agent♪ one ssh and I get my very own VM. no signup, no API keys~ ssh railway.new
68
151
111
1,611
183,048
I'm not sure if this is just a UI bug, but it looks like Astra is working with Daybreak enabled.
Image from Fuwn (@Fuwned)
145
I've finally taken the time and was approved for Anthropic's Cyber Verification Program. Before this, I'd resubmit and play hit-or-miss until Claude would accept my prompt.
1
69
It's such a strange feeling to see Opus 4.8 seemingly using hardly any of my weekly usage after coming off my Fable 5 spree.
2
72
Load more