high_byte

14.1K posts

high_byte banner
high_byte

high_byte

@high_byte

ai, software dev, cyber security and whatever else

Katılım Mart 2016
3.2K Takip Edilen5.7K Takipçiler
Sabitlenmiş Tweet
high_byte
high_byte@high_byte·
ever since the twitter algo dropped I keep getting comments on my pinned post. so if you are an LLM / AI / BOT or otherwise, please say so I can block you.
English
97
0
145
3.2K
Christy
Christy@REPOSTLOUNGEHP2·
I feel like giving this a try
English
172
955
10.1K
1.5M
high_byte retweetledi
“paula”
“paula”@paularambles·
this is wild
“paula” tweet media
Peter Steinberger 🦞@steipete

People freaking out over my AI spend. What nobody sees: Part of what excites me so much about working on OpenClaw is that I'm trying to answer the question: How would we build software in the future if tokens don't matter? We constant run ~100 codex in the cloud, reviewing every PR, every issue. If a fix on main lands, @clawsweeper will eventually find that 6 month old issue and close it with an exact reference. We run codex on every commit to review for security issues (as it's far too easy to miss). We run codex to de-duplicate issues and find clusters and send reports for the most pressing issues. We have agents that can recreate complex setups, spin up ephemeral crabbox.sh machines, log into e.g. Telegram, make a video and post before/after fix on the PR. There's codex that watch new issues and - if it fits our documented vision well, automatically create a PR of it. (that then another codex reviews) We have codex running that scans comments for spam and blocks people. We have codex instances running that verify performance benchmarks and report regressions into Discord. We have agents that listen on our meetings and proactively start work, e.g. create PRs when we discuss new features while we discuss them. We build clawpatch.ai to split all our projects into functional units to review and find bugs and regresssions. We do the same split for security with Vercel's deepsec and Codex Security to find regressions and vulnerabilities. All that automation allows us to run this project extremely lean.

English
50
80
2.4K
467.4K
high_byte
high_byte@high_byte·
@waitin4agi_ isn't seedance 2 like $1 per second? even at $0.5 it should be more than $1 per video, no? how is it so cheap for you
English
2
0
0
201
Varun Mayya
Varun Mayya@waitin4agi_·
8k USD spent on just Seedance credits for our next video. 14k generations, but we’ve selected just 30 shots. Regenerations again and again till we got it just right. Tomorrow :)
Varun Mayya tweet media
English
156
15
1.1K
79.7K
high_byte
high_byte@high_byte·
@javilopen I feel in the same hole. screaming into the void
English
0
0
1
9
Javi Lopez ⛩️
Javi Lopez ⛩️@javilopen·
F*ck, that felt good to get off my chest, xD
English
5
1
35
2.9K
Javi Lopez ⛩️
Javi Lopez ⛩️@javilopen·
I've gone from 50k-2M per post to literally a measly 3k per post. If Linus's reach is "worse", mine has literally been thrown straight into the shadowban bin 😂 NOBODY will see this. Well, yep, <3k people.
Javi Lopez ⛩️ tweet media
English
131
11
342
63.2K
high_byte
high_byte@high_byte·
@tibo_maker I paid for this sub, uploded my references, wrote a prompt, submitted - only to be hit with another paywall I need a more expensive subscription instantly wanted to unsubscribe only to find out I lose all my credits and access to the product - that I already paid for
English
0
0
0
42
Tibo
Tibo@tibo_maker·
genuinely don't understand the hype for higgsfield their new "Super Computer" seemed incredible, beautiful marketing tried it it created 2 seedance videos I asked it to clip them together got me to subscription screen 🤷‍♂️ it cost me 20% of my $49 subscription, so $10 the raw cost is about $1 so they resell this for x10 the price the value in Revid seems x10 better ✅ prompt it and it's done ✅ you don't have to clip sequences elsewhere ✅ full edit capabilities, a real video editor ✅ no scammy/shady pricing tactics
English
84
12
281
43.1K
high_byte
high_byte@high_byte·
@NickADobos here I am not reaching my $20/m cursor limits 🫠🫠🫠
English
0
0
1
110
high_byte
high_byte@high_byte·
we're already in nuclear war ai data centers are powered by energy from nuclear reactors ai race = nuclear war
English
0
0
3
339
high_byte
high_byte@high_byte·
@IntCyberDigest the requirements to trigger this are too much rewrite+set+question mark in rewrite rule+attacker knows the rewrite rules
English
0
0
4
651
International Cyber Digest
International Cyber Digest@IntCyberDigest·
‼️🚨 MAJOR IMPACT: AI just found an 18-year-old NGINX critical remote code execution vulnerability. It has been disclosed on GitHub including PoC code. - Affects NGINX 0.6.27 through 1.30.0 - Triggered via the rewrite and set directives in config - Update NGINX ASAP - NGINX is a widely used HTTP web server, be sure to check its prevalence in other products
International Cyber Digest tweet media
English
86
403
2.6K
935.4K
high_byte
high_byte@high_byte·
@hetmehtaa 18m? more like 18 of them. requires rewrite+set+have question mark+that the attacker knows these rules
English
0
0
1
692
Het Mehta
Het Mehta@hetmehtaa·
Every 3rd website you visit runs Nginx. 18,959,833 of them can be hijacked right now. A bug from 2008 just got a working exploit. CVE-2026-42945 (CVSS 9.2) No login. No access. Just one HTTP request. → Heap overflow → Worker process → RCE Patch ASAP to Nginx 1.31.0 or 1.30.1 PoC is already out: github.com/DepthFirstDisc…
English
34
121
810
143.8K
high_byte
high_byte@high_byte·
it looks like nano banana pro with prompt for 3x3 grid and then simply split the generated output evenly. at least that's how I replicated it
English
0
0
0
140
high_byte
high_byte@high_byte·
does anyone know how higgsfield shots work? it not only creates very high consistency, it actually returns separate images! not a single image grid. I want to build this as an API
high_byte tweet mediahigh_byte tweet media
English
1
0
0
180
high_byte
high_byte@high_byte·
@andreintg is there a technical limitation to run diffusion locally?
English
1
0
0
57
Andrea Intg.
Andrea Intg.@andreintg·
@high_byte yeah LCM is what we started with and it eventually matured into a custom c++ pipeline. the SDF evaluator is all local webgpu while the diffusion is being streamed from the server
English
1
0
3
74
Andrea Intg.
Andrea Intg.@andreintg·
I was a bit hesitant about showing stuff like this just a month ago 😅 When we started showcasing real-time AI + SDF sculpting, I was afraid professionals would laugh if I showed no effort on the input models. The shape strength slider was also hidden in our first iteration, so I had no choice but to at least try and knock some more interesting shapes together. Now that we're starting to focus on more powerful features and shape strength is finally unlocked, I'm starting to appreciate just playing with simple shapes. Different stages of production have different needs. Sometimes you want full authoring over your creations, while other times you just want to quickly explore new ideas.
English
175
410
4.4K
330.5K
high_byte
high_byte@high_byte·
@andreintg cool and is it LCM? is it running locally? in browser? that would be OP
English
1
0
0
72
Andrea Intg.
Andrea Intg.@andreintg·
@high_byte yeah you got it, for our intended users we're thinking less technical terms will come across as more intuitive. If we were to deploy this as a library, for comfyui for example, I agree we'd be better off using the technical names.
English
1
0
1
855
high_byte
high_byte@high_byte·
@C_3C_3 so basically: "to win this award you must put this genitalia in your mouth" which is an odd thing to say
English
0
0
1
377
C3
C3@C_3C_3·
To be eligible to win Best Picture at the Oscars… 1. At least 1 Non-White/Non-Straight lead or Significant Role. 2. At least 30% minor roles Non-White/Non-Straight people. 3. At least 2 Departments headed by Non-White/Non-Straight people. This is why Nolan did it. Coward.
C3 tweet media
English
3.3K
4.6K
39K
34.8M
Brian Roemmele
Brian Roemmele@BrianRoemmele·
It’s happening…
English
984
3.8K
22.7K
1.9M