Python developer, author of Make Art with Python. On a mission to bring art and play to software development. Previously lead DevRel at Datadog.

Joined March 2009
428 Photos and videos
Pinned Tweet
23 Sep 2024
I wrote about the generative video editor I've been working on for the past year, and the progress so far: makeartwithpython.com/blog/a…
1
4
24
2,831
Kirk Kaiser retweeted
If the knicks could pull that comeback then openai can come back against anthropic
27
29
647
40,116
Kirk Kaiser retweeted
Asked a question so dangerous they sent my ass to haiku
78
394
8,981
209,873
Kirk Kaiser retweeted
1
3
24
1,930
Wonder what percentage of market price discovery is being done by LLMs at this point and whether their sycophancy might be exploitable
70
Kirk Kaiser retweeted
n config options can often be replaced by k options (with k << n) where k just lets the user express what thing they want to optimize for (typically some tradeoff between latency, throughput, and cost). This is often very hard! But a 10x better user experience.
7
4
56
5,787
Kirk Kaiser retweeted
babe wake up, the next chapter of the deel espionage saga just dropped
25
27
572
225,838
Kirk Kaiser retweeted
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10… 🧵 pyfound.blogspot.com/2025/10…
148
190
1,035
480,573
Kirk Kaiser retweeted
There are a sophisticated set of enabling technologies, but the feature’s goal was incredibly simple: to capture the vitality of the world around us. The Live Photo patent has a picture of my daughter as a toddler spinning in a dress in front of our house. As with a Live Photo, the more time passes, the more endearing it’s become.
26 Oct 2025
live photos hit this perfect intersection of tech & tenderness. it might be the most quietly brilliant thing apple ever made. it’s a subtle wordless trick that turns stillness into something breathing. i just stumbled on one & it hit like a gut punch. ridiculous that all photos have a tiny slice of life before the pose, the laugh or blink right before the photo froze… it makes your chest ache hardcore esp for a memory you can never ever have again.
16
43
831
124,106
23 Oct 2025
Banger line up! If you’re anywhere near sf you should go
23 Oct 2025
I'm teaming up with @samuelcolvin to launch a new meetup series, Py AI. Our first one is November 11 in San Francisco with talks from @pydantic, @modal, @reductoai and @fastmcp. Hosted by our friends at @pebble_bed. Link to join below!
126
Kirk Kaiser retweeted
22 Oct 2025
Wrote this up as a blog post and decided that yes, for me it still matters vickiboykis.com/2025/10/20/i…
16 Oct 2025
Note I’ve been thinking about for a while
9
19
167
40,373
22 Oct 2025
okay the claude code questions are a really great feature
63
Kirk Kaiser retweeted
inkwell is a self-funded project. if you enjoy following it's development & would like to see it become a reality - please consider leaving a tip.
each npc is an ai persona that you can customise. they have context of the items, scenery, characters and actions that have taken place around them. they're still in the early stages, but this will evolve into controlling the actions they take as well.
5
7
45
3,442
21 Oct 2025
One of the arguments that has been used is that LLMs are “transformative” and thus cannot be considered derivative on their training data. But in the visual domain we can very much see that these large models are in fact derivative from what they were trained on.
OpenAI has again tightened copyright restrictions on Sora after Breaking Bad star Bryan Cranston saw himself in Sora 2 generations during the launch and contacted SAG-AFTRA. They have released a joint statement.
1
2
437
Kirk Kaiser retweeted
simple
Replying to @DFinsterwalder
Important info. The issue in that benchmark seems to be ollama. Native llama.cpp works much better. Not sure how ollama can fail so hard to wrap llama.cpp. The lesson: Don’t use ollama. Espacially not for benchmarks.
12
26
348
29,685
It’s such a good hack to ask if you’re accidentally being boolean in decision making. Humans have a tendency to think in all or none, but you almost always eliminate all or none decisions and find a better blend between options.
A mistake I’ve made a bunch of times is to focus on the bottleneck X of a system for 6 months until it’s no longer a bottleneck. And then focus on the next bottleneck Y for 6 months. What I should have done is to anticipate Y and focus on both X and Y in parallel.
112
I wonder what the Sora 2 release does to the penalties from the NY Times lawsuit
1
194
Does this change the math on whether or not LLMs are really as derivative as we think they are?
95
28 Sep 2025
I’ve done a lot of things conventionally seen as difficult but man they don’t even register as blip relative to taking care of an aging relative
2
104