Max Caldwell
@growth.wtf
Bootstrapping a software company. In my free time I try to do things with data and statistics, or post random thoughts here.
Pinned
I was wrong. OpenAI should have fired Sam.
August 16, 2025 at 5:29 PM
I was wrong. OpenAI should have fired Sam.
Me after my 7th cup of coffee in the morning
In Super Mario 64 DS, by alternating pressing Y+Up/Down/nothing/Down/nothing every 5 frames, Mario can vibrate up every slope in the game.
July 27, 2025 at 4:11 PM
Me after my 7th cup of coffee in the morning
openai.com/index/gpt-4-1/ and it's dead
April 14, 2025 at 9:08 PM
openai.com/index/gpt-4-1/ and it's dead
Gpt-4.5 has a fascination with the word " clearly" but I've never seen a model have before.
March 9, 2025 at 6:02 PM
Gpt-4.5 has a fascination with the word " clearly" but I've never seen a model have before.
Hello new followers! If you are seeing this, introduce yourself!
February 28, 2025 at 9:37 PM
Hello new followers! If you are seeing this, introduce yourself!
Just heard about this journal on the video game history hour podcast! Seems like an awesome project.
Hello Bluesky! We're thrilled to introduce ourselves as ROMchip: A Journal of Game Histories. We develop, edit, & publish ad-free, open access game history research for a range of audiences. We're anti-paywall, anti-gatekeeping, reader funded. Read, Like, Subscribe.
romchip.org
romchip.org
ROMchip
ROMchip: A Journal of Game Histories is a platform designed for the advancement of critical historical studies of games.
romchip.org
February 28, 2025 at 1:19 AM
Just heard about this journal on the video game history hour podcast! Seems like an awesome project.
Gpt-4.5: The ultimate pick-me.
Seriously feels like they just decided to YOLO a checkpoint because Claude 3.7 came out.
Seriously feels like they just decided to YOLO a checkpoint because Claude 3.7 came out.
February 27, 2025 at 8:14 PM
Gpt-4.5: The ultimate pick-me.
Seriously feels like they just decided to YOLO a checkpoint because Claude 3.7 came out.
Seriously feels like they just decided to YOLO a checkpoint because Claude 3.7 came out.
I think these people are all in love with the robot. Like a romantic love.
Ok, eq and creative intelligence was a focus for them this time? So I see why they're starting with that... Not 100% sure why they're focusing on it though.
Gpt-4.5 demo: can help you if you have absolutely no social awareness? Very weird place to start the demo
February 27, 2025 at 8:10 PM
I think these people are all in love with the robot. Like a romantic love.
Ok, eq and creative intelligence was a focus for them this time? So I see why they're starting with that... Not 100% sure why they're focusing on it though.
Gpt-4.5 demo: can help you if you have absolutely no social awareness? Very weird place to start the demo
February 27, 2025 at 8:06 PM
Ok, eq and creative intelligence was a focus for them this time? So I see why they're starting with that... Not 100% sure why they're focusing on it though.
Gpt-4.5 demo: can help you if you have absolutely no social awareness? Very weird place to start the demo
February 27, 2025 at 8:03 PM
Gpt-4.5 demo: can help you if you have absolutely no social awareness? Very weird place to start the demo
Reposted by Max Caldwell
I can't believe someone just decided 0! = 1 and we LET THEM GET AWAY WITH IT
February 23, 2025 at 7:42 PM
I can't believe someone just decided 0! = 1 and we LET THEM GET AWAY WITH IT
Tom Scott and Adam scott are cousins??!?!
February 23, 2025 at 3:30 AM
Tom Scott and Adam scott are cousins??!?!
Happy, sad, or some secret third emotion.
February 23, 2025 at 12:21 AM
Happy, sad, or some secret third emotion.
Hacker News comments are truly some of the most insane and entitled tech takes
February 22, 2025 at 9:47 PM
Hacker News comments are truly some of the most insane and entitled tech takes
This is 100% correct
one of my math colleagues is increasingly convinced that AI/ML people are using the word "manifold" in ways designed specifically to troll him
February 21, 2025 at 4:05 AM
This is 100% correct
There is so much dogshit software in the world
February 21, 2025 at 3:54 AM
There is so much dogshit software in the world
Forget Google Meet.... Introducing—Google Feet!
February 14, 2025 at 5:58 PM
Forget Google Meet.... Introducing—Google Feet!
caffeine pill and coconut water: breakfast of champions
February 7, 2025 at 5:21 PM
caffeine pill and coconut water: breakfast of champions
Recommending my friend who is building on 4o-mini check this out
Three new Gemini models today: Gemini 2.0 Pro (Experimental), Gemini 2.0 Flash and Gemini 2.0 Flash-Lite
Flash is cheaper than GPT-4o mini, and Flash-Lite is HALF the price of OpenAI's cheapest model!
Updated llm-gemini plugin adding support for the 3 new models: simonwillison.net/2025/Feb/5/g...
Flash is cheaper than GPT-4o mini, and Flash-Lite is HALF the price of OpenAI's cheapest model!
Updated llm-gemini plugin adding support for the 3 new models: simonwillison.net/2025/Feb/5/g...
Gemini 2.0 is now available to everyone
Big new Gemini 2.0 releases today: - **Gemini 2.0 Pro (Experimental)** is Google's "best model yet for coding performance and complex prompts" - currently available as a free preview. - …
simonwillison.net
February 5, 2025 at 5:24 PM
Recommending my friend who is building on 4o-mini check this out
@timkellogg.me Good job getting to the front page of hacker News
February 5, 2025 at 4:54 PM
@timkellogg.me Good job getting to the front page of hacker News
This is going to sound crazy but I think "energy field" people who go based on vibes are on to something.
I think they've created a sort of mathematical construct in their head + mistakenly imputing it onto reality, but modeling competence, positivity, or other attributes as or networks makes sense
I think they've created a sort of mathematical construct in their head + mistakenly imputing it onto reality, but modeling competence, positivity, or other attributes as or networks makes sense
February 2, 2025 at 7:55 PM
This is going to sound crazy but I think "energy field" people who go based on vibes are on to something.
I think they've created a sort of mathematical construct in their head + mistakenly imputing it onto reality, but modeling competence, positivity, or other attributes as or networks makes sense
I think they've created a sort of mathematical construct in their head + mistakenly imputing it onto reality, but modeling competence, positivity, or other attributes as or networks makes sense
O3 materially moves up my timeline of ASI/agi
February 1, 2025 at 1:48 AM
O3 materially moves up my timeline of ASI/agi
Holy shit, just got o3 access
January 31, 2025 at 8:27 PM
Holy shit, just got o3 access
I'm really tired right now but this Ray dalio book that's coming out is complete cherry picked bullshit right?
I mean I respected the guy in his team did a lot of research. But it kind of sounds like if you squint data long enough you Will find a pattern.
I mean I respected the guy in his team did a lot of research. But it kind of sounds like if you squint data long enough you Will find a pattern.
January 30, 2025 at 4:35 AM
I'm really tired right now but this Ray dalio book that's coming out is complete cherry picked bullshit right?
I mean I respected the guy in his team did a lot of research. But it kind of sounds like if you squint data long enough you Will find a pattern.
I mean I respected the guy in his team did a lot of research. But it kind of sounds like if you squint data long enough you Will find a pattern.
It's very likely that humans will be convinced AI is sentient well before it actually happens.
January 25, 2025 at 4:44 PM
It's very likely that humans will be convinced AI is sentient well before it actually happens.