creator of aftermatter corp. (github.com/aftermatter), shrike ⚡︎, contextplus (2k) and cargo install safeinstall. forloop@poke.com
- Tweets 27,636
- Following 2,092
- Followers 9,695
- Likes 70,812
Pinned Tweet
Sure! Here's how the tightest (known) packing of 67, 6767 and 676767 squares in a larger square looks like:
I made the most optimized algorithm to find out the tightest packing of n smaller squares in a larger square
test it out: forloopcodes.github.io/least… (runs on your gpu)
more details:
input: n (number of squares)
intermediate: s (side length)
output: [[x, y, angle]*n] (array of position of each square)
solver (squarepack): pack(n) returns (s, [[x, y, angle], …])
scale: n=100000 in about 4 s, 0.08 s in browser, n=1000001 in 0.2 s.
verifier: p_ij = ½ + (|cos φ| + |sin φ|)/2 − max_a |a·(c_i − c_j)|, φ = θ_j − θ_i
flow:
1. best of all in O(n): Grid ⌈√n⌉, Göbel strip/square families, L-extensions, the √7 family, DeVincentis,Wainwright 19, Schadt 50/171
2. cache lookup
3. tilted-block search: p×q block rotated 45°, remaining space filled by walls anchoring
4. numerical compaction: Penalty energy E = Σ max(0,penetration)² + Σ max(0,protrusion)², analytic gradient, L-BFGS, shrink s by bisection, basin-hopping.
github: github.com/forloopcodes/leas…
I'm sure you always wanted to know how the tightest (known) packing of "n" squares within a larger square looks like... 😍
2
1
15
5,188
OpenAI has nerfed chatgpt for everyone, the way tibo and altman used to hype up their models 2 months ago and gave resets, its all worth nothing, they replaced the model served on their website with the dumbest model they ever had
if i was on a free plan i'd prefer to use gemini instead, atleast it would give more accurate answers than chatgpt
The way they were hyping astra just to get mogged by a mythos distilled model and then paused 200$ subscriptions mentioning they didnt have any compute, then be more greedy and change 5x plan's limits to "more usage"
they're going to introduce a 500$ plan where they would charge more for ultrafast for an actually cheaper cerebras inference
this is literally: if you don't pay for it, you get a 2b model
also, they often hype up free users with free plus trials, when you try entering your card, rejected, this card has no issues at all while paying for claude sub. the card just goes through, not just one, i have tried many cards over many accounts
what a pathetic situation of a company which was unbeatable 2 months ago
no hate for openai, but they have come to a point that if they dont bring up a comeback at devday it would be impossible to catch up with anthropic
1
4
101
forloop retweeted
Full desktop Google Chrome, Blender, and more, on Android?
Yes, please!
18
9
257
11,759
i swear if they make sonnet cheaper then anthropic will win so hard openai could never catch up
1
16
502
we HAVE to nuke SF
i’m looking for a VC girlfriend. having talked to 100s of VCs and founders, here’s why i’m 100% serious that this is a win-win:
- i’m raising my series A for @KoyalAI in a few months. If you backchannel me, we’ll both get to Series B faster.
- i’ll be your scout for founders before they get to YC/a16z. My DMs are full of ambitious people starting companies.
- i can take you to nice restaurants and you can expense it as “diligence.”
- you probably take better care of yourself than most founders. You can womansplain biohacking to me
- most of my friends are founders who hate VCs, so you’ll get a front-row seat to the gossip. I expect the same privileges in return.
- we’ll never have to argue about who pays for coffee chats. Your fund does.
- if the relationship works out, great. If it doesn’t, we still generated deal flow for our respective organizations.
- we probably won’t get bored talking about tech, but can also go to burning man.
please DM or tag relevant friends.
23+ years old, seed and above, serious inquiries only, thx.
again, not a joke. X do your thing.
4
72
3,024
where are all the codex praisers? ive not seen one (1) person who praised codex and openai after the opus 5.5 launch
are people even using their models now? did anthropic already win the AI race?
6 Luna 的数据出来了。我不知道说 OpenAI 啥了，我就觉得犯恶心，你们自己看吧，现在是 Codex 用户不被当人了🤡
出之前大肆宣扬性能大幅提升、廉价到无法计量；出的时候价格降一半，说这同样会提高使用限制。
结果我一统计，额度从 75 亿砍到 44 亿，直接削了 40%🤡🤡一刀砍这么狠的真是头一次见😀只能说希望是我算错了
3
17
1,031
You can't ever call opus 5.5 "cheap". it is NOT a cheap model, its cost efficient but not cheap
Cheapest AI models right now:
1️⃣ GPT-6 Luna: $0.10 / $0.50
2️⃣ Qwen3.8 Flash: $0.15 / $0.47
3️⃣ DeepSeek V4.1 Flash: $0.15 / $0.60
4️⃣ GLM-5.3 Flash: $0.15 / $0.50
5️⃣ Gemini 3.8 Flash: $0.75 / $3.75
6️⃣ GPT-6 Sol: $2.00 / $10.00
7️⃣ Claude Sonnet 5: $2.00 / $10.00
8️⃣ Grok 4.7: $2.00 / $6.00
9️⃣ Qwen3.8 Max: $2.00 / $6.00
🔟 Claude Opus 5.5: $4.00 / $20.00
The price war is real.
4
122
5,725
forloop retweeted
you either die a claude code chud or you live long enough to see yourself build a harness
22
6
429
12,765
tonight my claude max 20x will end, the pro plan would be such a pain to use with limits AAAAAHAAGGGGGGGGGHHHHHH
3
26
897
Someone
at what point of life would you require 1000 sources to check if tung tung tung sahur is real or not
Introducing Agent Ultra - a step change in deep research
Agent Ultra orchestrates swarms of agents to perform exhaustive research, build comprehensive lists, and answer questions requiring thousands of sources.
In both evals and vibes, it's state of the art
1
18
807
I've been building something since many days and would love any opencode employee to try it in alpha.
is there anyone who can get me in one's dms
3
25
7,199
gpt 5.6 luna vs claude haiku 4.5:
as a claude user i have more iq than an average chatgpt user
imagine your model losing to a 1 year old model in basic text analysis for free users, at this point you're actually misleading free user's knowledge, even gemini flash lite would outperform luna at this point
4
24
1,384