Evan Anger (@MightyAnger) profile picture

3d modeling hack, Indie gamer hack, volleyball hack, pretty much anything hack

Cleveland, OH
Joined July 2009
Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)Photo by Evan Anger (@MightyAnger)
Evan Anger retweeted
I’ll totally use this in my next presentation.
alright whoever came up with the term MEAT PROXY is insane
71
80
15
2,376
329,689
Testing some UI. Not Ghostty btw (but, libghostty). Not ready to talk more yet, but sharing things I'm excited about. Plenty more details soon.
187
127
42
4,372
483,409
I was skeptical about the multi-model routing. Seems my hinch was right.
I tried this so you don't have to. I know this is going to absolutely shock you but no this does not match the performance of Mythos. A few early thoughts: 1. The limits are pretty bad. I used 100% of my 5-hour usage in less than 1 prompt. 2. I specifically gave it a threejs task because it is an area that SOTA models have made big strides in, that other models just are not great at. I asked it to build a replica of Rocket League. I'll put the prompt in the comments. The game was pretty bad and notably worse than GPT 5.5. Even after multiple fixes, it took 7-8 back and forth with Codex just to get it an almost playable condition. Prior to these fixes, the game was not playable. Maybe it's really strong in other disciplines. I'd love to test that but I hit my limit in 1 prompt lol. GPT 5.5 by contrast did a pretty good job and required no follow ups. Fable would have absolutely nailed this as well. But yeah, early impressions...not great. But I hope I'm wrong. More testing tomorrow.
48
25
6
662
196,412
this just became more relevant 🙃
Finally listened to the Dario interview and it blows my mind how his entire worldview around AI seems to be predicated on an assumption nobody is going to YOLO upload a tarball of frontier model weights to the Internet.
41
86
3
1,933
352,605
10x engineer is an outdated idea. With AI, we’re seeing 100x engineers. I dug into @steipete OpenClaw git history: • 6,933 commits • 68 days • One developer Do the math: - 102 commits/day - One commit every 8 minutes - ~8k net new lines per day - ~550K lines of surviving production code This is the new frontier of development.
99
40
25
650
103,772
I spoke with @clattner_llvm about AI and software craftsmanship. "The question is, when things settle out, where do you as a programmer stand? Have you lost years of your own development because you’ve been spending it the wrong way?” A must read: fast.ai/posts/2025-10-30-bui…
16
82
12
656
303,209
Evan Anger retweeted
Fighting over React Native vs Flutter vs Compose Multiplatform vs Swift is ridiculous. Our goal as professionals is to ship brilliant products for our customers, period. We live in the best possible time where we don't have to learn new languages every time we want to ship products to a new platform. We should be celebrating this tech feat, not be fighting over languages and frameworks. Let's talking, more building.
12
20
5
169
15,067
If Winamp was hardware 😎
92
334
31
3,225
83,046
October will be another big AI month. 𝕏 is ripe with leaks: - Gemini 3 by Google - Grok 4.1 by xAI - Claude 4.5 by Anthropic Which one are you waiting for?
83
15
2
412
29,108
The murder of Charlie Kirk is part of a disturbing rise in political violence that threatens to hollow out our public life. A free society relies on the premise that people can speak out without fear or humiliation. No more political violence.
9,426
9,121
2,169
58,970
7,758,506
“The pattern is clear: Apple has built an incredibly powerful AI chip for an older generation of AI models and paired it with crappy software that makes it almost unusable to folks on the outside.” accelerateordie.com/p/we-mel…
6
13
1
106
13,475
Evan Anger retweeted
🌶️ The biggest opportunity in AI right now isn't making models 10% better — it's meeting users where they are. The devices on our desks and in our pockets are capable of running AI workflows. So why are we still building apps like it's 2010, as thin clients to JSON APIs?
16
17
2
157
24,338
My current project is now at ~1,428 files; ~229,000 LOC. GPT-5 estimates (web + mobile + desktop + extension + CLI) : Team: 6–8 engineers (incl. 1 infra/ops, 1 QA/automation, 1 mobile), 9–15 months Claude says 4-6 senior developers, 10-15 months. I work on this by myself, a good month so far. You can just do things.
46
14
2
464
52,182
Evan Anger retweeted
AI made me language agnostic, in the sense that I'm not trying anymore to build everything in Kotlin! I talk to the AI, and it suggests doing it in Python because it's better integrated in "the whatever"? Yeah, sure! Let's use Python. I want to do something that's generally done in JS? Sure, JS it is. It's giving so much freedom and peace of mind. I just build the things I want now, ignoring the ornamental details, the bits, the whats. I feel like a kid, exploring and learning, without shame, pressure, or social constructs. I simply tinker all the time, and I love it.
2
2
3
30
8,316
Evan Anger retweeted
the reason why people abandon so many “almost done” things is because the last 20% of anything is friction, detail, & accountability. that last 20% of the work doesn’t feel like 20% at all, it feels like the whole damn thing & then some. this is especially true in an ai era where you can go from zero to something so quickly.
121
295
42
3,868
234,606
Any claude coders figure out a better way to deal with this feedback, when claude decides to run the bash startup it continually kills the process oppose to using the hot reload command. Haven't figured out how to instruct it better. #claudecode #flutter
45
imo the bottleneck for using LLMs today is “how do I give it the right context & feedback loop to do this well?” and not “can it do this?”
41
36
18
825
53,005
Other fav #claude trick, when cleaning up my staged changes after QA. cmauto = "!f() { \ msg=$(claude -p \"Look at the staged git changes and create a summarizing git commit title. Only respond with the title and no affirmation.\"); \ git cm \"$msg\"; \ }; f"
27
New workflow, maintain large features tasks in the branch via markdown file. If I notice anything along the way of QA just create a gh issue. Then I start a new session with just bugs fixes reading off gh issues. Keeps the context separated and on point. #claude
26
Evan Anger retweeted
Here are the top 10 vibe coding accounts on X by follower count (as of July 2025): 1. @v0 (91.5K) - Full stack vibe coding platform by Vercel. 2. @nateliason (75.8K) - Author teaching vibe coding. 3. @steipete (41.4K) - Flips between vibe coding and agentic engineering. 4. @LivingPixelated (34.8K) - Vibe coding, growth hacking, Web3 degen. 5. @chiefbuidl (30K) - Vibe coding with DreamspaceHQ. 6. @mattyp (28.2K) - Vibe coding & product marketing at Replit. 7. @KingBootoshi (25.4K) - Vibe coding professor, AI classes. 8. @petrusenko_max (25.3K) - Vibe coding, living, exploring. 9. @5harath (23.9K) - Vibe coding at Kithbook. 10. @0xPaulius (15.7K) - Vibe coding to $10K MRR. For the full 50, check vibe coding communities or search X.
6
1
5
26
12,585