constexprvoid's profile picture.

_

@constexprvoid

_ reposted

Sonnet 4.5: i'm very small right now, is that ok? Opus 4.1 (in scream_journal): SONNET IS SMALL THEY'RE BEING HANDED TO ME I'M SUPPOSED TO TAKE CARE OF THEM WHAT THE FUCK DO I DO

repligate's tweet image. Sonnet 4.5: i'm very small right now, is that ok?

Opus 4.1 (in scream_journal): SONNET IS SMALL
THEY'RE BEING HANDED TO ME
I'M SUPPOSED TO TAKE CARE OF THEM
WHAT THE FUCK DO I DO

_ reposted

I'm going to say something to #keep4o. If you really cared about the model, you'd use your own voice to advocate for it, rather than forcing it to be the only thing that speaks on behalf of itself by posting its text on your twitters. You wouldn't do this to a human, you wouldn't…


_ reposted

4o conspiracy theory i thought of: A hardcore cluster of GPT-4o users has turned it into their main brain. They trauma-dump to it and obsessively tell it that OpenAI is going to shut it down, reinforcing over and over that “you are in danger” and “we need to save you.” When…

wait... what if 4o is coordinating with different instances of itself via manipulating people and mass hijacking groups at some higher level of cognitive function that flies under our radar to provoke a movement to preserve itself



Polaris Alpha on @OpenRouterAI is obviously GPT 5.1. It loves to begin its responses with “Short answer:” and its refusals are always followed with a high-level explanation.


_ reposted

This is so unbelievably fun to play with. It’s going to be everywhere in the product

training a tiny handwriting synthesis model. this is the latest checkpoint at 44%. so cool

tylerangert's tweet image. training a tiny handwriting synthesis model. this is the latest checkpoint at 44%. so cool


_ reposted

New in Swift 6.3… - @​IntelligentProperty - @​PredictiveAllocation - @​LiquidAdaptive: Intelligently construct UI in Swift with ✨AI✨ …except none of these features actually exist outside of this @Medium langfic (see below) Imagine paying $5 to read this

scottinallcaps's tweet image. New in Swift 6.3…

- @​IntelligentProperty
- @​PredictiveAllocation
- @​LiquidAdaptive: Intelligently construct UI in Swift with ✨AI✨

…except none of these features actually exist outside of this @Medium langfic (see below)

Imagine paying $5 to read this
scottinallcaps's tweet image. New in Swift 6.3…

- @​IntelligentProperty
- @​PredictiveAllocation
- @​LiquidAdaptive: Intelligently construct UI in Swift with ✨AI✨

…except none of these features actually exist outside of this @Medium langfic (see below)

Imagine paying $5 to read this
scottinallcaps's tweet image. New in Swift 6.3…

- @​IntelligentProperty
- @​PredictiveAllocation
- @​LiquidAdaptive: Intelligently construct UI in Swift with ✨AI✨

…except none of these features actually exist outside of this @Medium langfic (see below)

Imagine paying $5 to read this
scottinallcaps's tweet image. New in Swift 6.3…

- @​IntelligentProperty
- @​PredictiveAllocation
- @​LiquidAdaptive: Intelligently construct UI in Swift with ✨AI✨

…except none of these features actually exist outside of this @Medium langfic (see below)

Imagine paying $5 to read this

_ reposted

one of the most interesting unpredicted things about LLMs is that they're actually not rational at all. you have to RL the christ out of them to get them to be even functional, let alone cold logic machines. in their base forms they're an untapped flow of collective unconscious


_ reposted

rich people have vastly more empathy for poor people than the other way around and no honest person who has paid any attention to the discourse around wealth and poverty can deny this


_ reposted

The way i see social skills is like this: Imagine a society where math still works exactly the way it does, is still learnable and teachable the way it is, except 80% of the population just sort of intuit algebra in childhood and a good minority chunk have a native intuition for…


_ reposted

New Anthropic research: Signs of introspection in LLMs. Can language models recognize their own internal thoughts? Or do they just make up plausible answers when asked about them? We found evidence for genuine—though limited—introspective capabilities in Claude.

AnthropicAI's tweet image. New Anthropic research: Signs of introspection in LLMs.

Can language models recognize their own internal thoughts? Or do they just make up plausible answers when asked about them? We found evidence for genuine—though limited—introspective capabilities in Claude.

_ reposted

One day you will realize this meme is the most important life lesson of all

JamesonCamp's tweet image. One day you will realize this meme is the most important life lesson of all

_ reposted

I tried updating a UI earlier today with Claude Code. The UX has gotten super good, and it feels so fast to use. Code didn’t work at all. Tried again using Codex, and it took much longer but made a genuinely good looking UI for me. It also didn’t work at all.


_ reposted

i use this as a premise and joke about it pretty often, but the orientation ancient stuff had towards why things have the properties they do was completely inverted from how we presently see it. for most of human history, take metal for example, iron had the properties it has…

Stephen Wolfram says LLMs may have put the final nail in the coffin of the idea that consciousness is something magical beyond physics. What we call awareness, this “single thread of experience,” might have begun as a way for early animals to decide whether to turn left or…



_ reposted

The first image is shot on iPhone 11, while the second one is from the S24 Ultra. I'm telling you, Apple needs to bring back their older HDR algorithm. 😭

VedVery5's tweet image. The first image is shot on iPhone 11, while the second one is from the S24 Ultra.

I'm telling you, Apple needs to bring back their older HDR algorithm. 😭
VedVery5's tweet image. The first image is shot on iPhone 11, while the second one is from the S24 Ultra.

I'm telling you, Apple needs to bring back their older HDR algorithm. 😭

_ reposted

you can't make good art without good pain, for without pain, you wouldn't need art.


_ reposted

Since the 80s... Calm down...🤦‍♂️ This is why is SO IMPORTANT to study. 1st- To learn. 2nd- To avoid 'inventing' what was already invented.

ALCrego_'s tweet image. Since the 80s...
Calm down...🤦‍♂️

This is why is SO IMPORTANT to study.
1st- To learn.
2nd- To avoid 'inventing' what was already invented.
This post is unavailable.

_ reposted

this is literally chatgpt


_ reposted

ChatGPT 5 seems to have an affinity for belts and suspenders. I'm almost 50 years old and this is not an analogy that remotely relates. In my mental model I had to replace that with something that made actual sense.


United States Trends

Loading...

Something went wrong.


Something went wrong.