David Cole
@irondavy
Followers
4K
Following
20K
Media
568
Statuses
7K
Design at @castlexyz. Making distinctions.
San Jose, CA
Joined June 2008
🌋 After years of being an internet recluse, I’m now bursting with announcements:. 1. @castlexyz has been my full focus for >5 years now, and I’m proud to share we recently closed a $15M round led by @khoslaventures to keep pursuing the dream of HyperCard, Flash, MarioPaint et al
16
20
377
Wow no way
0
0
1
What are some beginner pieces of information for somebody just getting into knowledge.
0
0
8
Very special chance to work in gen AI that isn’t vapid or anti-humane, on a product that’s already working made by a team filled with heart 👇.
I’m hiring a product designer for Sudowrite. This is a dream job for a designer who loves to write. The beginning of the listing is… a mini therapy sesh for me. You’d think it’d be chill after that. But no, it goes on to talk about our company cult….
1
2
7
The full cycle goes like this: .1. What looks good in a keynote gets shipped first.2. Over the beta they fix the glaring accessibility issues, but doing so dampens how slick the design system is.3. In 1-2 OS releases they find a reasonable middle ground and we forget about this.
3
6
90
I’ve just published the most ambitious piece of writing I’ve ever done, almost 20k words on Buster Keaton, Tom Cruise, and a hell of a lot of trains: .
2
4
19
Well now isn’t this just the cutest lil book of tanks you ever did see 🥺
0
0
4
This book was my favorite discovery from digging through this stuff, how crazy cool are these?!
Care to admire a bunch of gorgeous cross-section illustrations by the master Stephen Biesty? After learning he passed just last year, I combed over his all his books and wrote a brief retrospective:
0
6
6
Care to admire a bunch of gorgeous cross-section illustrations by the master Stephen Biesty? After learning he passed just last year, I combed over his all his books and wrote a brief retrospective:
0
6
33
Just today I watched ChatGPT start hallucinating so consistently that I had to jump to a new thread. where it immediately referenced the hallucination, I assume because of this feature. Given the tendency for an LLM thread to decay over time, isn't this just poisoning the well?.
we have greatly improved memory in chatgpt--it can now reference all your past conversations!. this is a surprisingly great feature imo, and it points at something we are excited about: ai systems that get to know you over your life, and become extremely useful and personalized.
5
1
11
RT @millsbaker: I really enjoyed a post by @joulee, occasioned by her convos with @henrymodis, on design in the LLM era. It was posted on a….
0
1
0
New post by me on Dead Horses, on the role of words in an LLM future
0
0
4
5. As part of this launch we’ve rebranded our older publication to RATS FROM ROCKS which will focus on culture and spirituality, where you can find my recent piece on the moment of awakening, and some podcasts I did with Mills on mysticism
0
0
22
Mills just dropped a banger that could’ve been titled FUCK “BEST PRACTICES” and in a couple days I will publish our first in a series of podcasts where we close-read NOTES ON THE SYNTHESIS OF FORM by Christopher Alexander
3
5
41