micah (@micahomnd) header
micah (@micahomnd) profile picture

swe @ stealth (@hf0 ‘25), cs @uwaterloo

San Francisco, CA
Joined September 2023
Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)Photo by micah (@micahomnd)
Lilian Weng @ OpenAI on Harness Engineering a deep breakdown of harness design patterns, workflow automation, persistent memory via file systems, and how modern agents like claude code actually work. the best blog post I've read all month. link: lilianweng.github.io/posts/2…
18
125
4
1,482
118,840
moved to sf expecting the apocalypse based on how people talk about it online but if you avoid like 2-3 specific neighborhoods, it's legitimately one of the best cities on earth don't listen to the haters
4
37
1,488
I'm convinced Chinese GitHub repositories and papers (Kimi Delta Attention, etc.) will increase your rate of learning by ~3x, esp. csdiy.wiki
12
229
6,206
I come back to this blog anytime I hit a cognitive wall when learning something
95
992
50
13,513
304,302
Sometimes I wish I went to Stanford they have a whole course where you literally build an LLM from scratch (tokenization, custom kernels, distributed training, alignment, etc.) this kind of exposure as an undergrad is so nice
22
35
5
969
47,542
the average person in SF is not average the girl you have a crush on is doing a PhD at Stanford and research at a frontier lab, the nerdy guy in the cafe is doing 10m+ ARR
1
13
894
it's crazy that future cs students won't know what stackoverflow and leetcode are
3
1
15
1,154
This is one month of Georgia Tech's CS8803 deep, paper-backed breakdowns of embeddings, reasoning/alignment, and agent harnesses/long context insane value, and the resources are all public: cocoxu.github.io/CS8803-LLM-…
3
78
1
665
20,799
called it a "bug" instead of "edge case trajectory drift" and they kicked me out of SF
9
770
Spotted in SF: a VC telling a homeless guy he's "cash poor, but rich on paper" like a "typical founder" He added: "You're just not liquid yet, why don't you sell secondaries?"
8
895
Waterloo makes you write your exam here btw..
this is how midterm marks are released at waterloo
77
79
69
4,159
267,173
just learn from elite company blogs: - google for systems (gfs, mapreduce, etc.) - anthropic/openai for agents (lilian weng's blog) - stripe's blog for high-stakes engineering (zero downtime, idempotency, payments, etc.) it's absurd how much knowledge is free
16
85
3
1,441
54,257
software is entirely about doing the *right* work
1
1
12
873
if you're 18-25, in software/cs, and relatively ambitious, rn is the best moment in history to be alive
31
44
10
1,155
51,963
coded by hand today just to feel something
5
29
1,449
AI widened the skill gap in the most skill gapped industry there is lol 10x engineers are now 100x engineers 1x engineers are now 0.1x engineers
2
1
12
1,017
As someone who grew up in Canada, SF is easily the most beautiful city I've been to, by far
5
1
74
2,391
the levels of modern AI engineering: L1: basic prompt engineering, RAG L2: context window arch., dynamic pruning L3: harness engineering & loops L4: deterministic evals, synthetic data L5: "hey, are you conscious?"
3
7
596
CMU's new AI agents course is as good as it gets, and totally free it goes all the way from multi-agent coordination to computer-use agents, deep research, and advanced reinforcement learning algorithms incredible for anyone building agent systems
13
92
1
825
21,828
POV: a day as a human in 2027 7am: wake up 8am: AI solves 91 year old unsolved problem 8:07am: "it's fine, my career requires taste." 8:09am: open favourite model, start prompting 8:10am: notification, "taste v1.0" added in new model 8:15am: log in for UBI, go enjoy rest of day
4
511
The modern engineer's dilemma: - optimize for output = use AI - optimize for learning = don't use AI What's the most efficient way to optimize for both learning and output at the same time?
3
1
8
863