HyperL0gi

Exact same thing here.

Both evals and Human pairwise tests for our use case are giving Haiku 4.5 first place in pretty much all tests.

No we'll try understand if we need to change our prompts to match performance ...

edit: maybe this will help: https://platform.claude.com/docs/en/build-with-claude/prompt...

Alcohol consumption was already in decline before the massive increase in GLP-1 drug use. As someone who used Mounjaro, I think that's only the tip of the iceberg.

When on these drugs, you simply have no desire for such things, and I'm a beer and wine lover. Now that I've stopped, I'm slowly drinking again, but I feel my taste has changed.

I don't know if it was the drugs or the time that passed, but I simply can't drink or have the same desire as before (it could be my age, too, lol).

Since you conveniently stopped at the beginning of the story, let me finish it for the sake of clarity.

In 2018, the president was Michel Temer, not Bolsonaro. Temer had been Dilma Rousseff's vice president and took office after she was impeached in 2016. Dilma, by the way, was Lula's handpicked successor. That December, Temer signed Law 13.756, which legalized sports betting but left the regulation for later.

That regulation came under Lula: on December 29, 2023, he signed Law 14.790, which set a licensing fee of up to R$ 30 million and authorized betting companies to operate legally in Brazil.

So, technically speaking, it started with a president elected on Dilma's ticket and was sealed by Lula.

Recipe for winning: introduce a problem, watch society get destroyed by it while you profit from it, then when a big portion of the poor population has too much debt because of the problem you legalized, ban it and offer a solution a couple of months before the election. Save the people, get elected. Win win

I think specs are incredibly valuable for who's writing/developing the software as a thought exercise, and pretty much useless for anyone else, especially after you ship it as it'll always get outdated very quickly.

GPT 5.6 Luna is ten times cheaper than GPT 4o and outperforms it. I genuinely don’t think Luna is more expensive to serve than GPT 4o.

If you were charging $10 per month per user for your SaaS when GPT 4o was launched, you can easily charge $5 now and increase your profit.

I’m not sure why you believe things will become more expensive.

Is anyone using DSv4 for their agents that is not related to writing code? Curious about use cases specially for someone using gpt-5.4 mini for classification, categorization, etc

I always admired people who spent hours configuring their local setup, but this was never me. For me I was always looking for the one script setup that would just get me 80% home and that was good enough. I’ve never had the patience to configure my vim or tmux bindings so I’d always use what out there. It’s not that I think it doesn’t matter but I could never find the motivation to spend that amount of time with it. I guess I always preferred to get from someone who already spent the time on. Open source is beautiful

My point is why the sudden change in tone? I’m not dismissing the models’ capabilities.

- https://www.anthropic.com/research/glasswing-initial-update

- https://www.axios.com/2026/04/08/anthropic-mythos-model-ai-c...

- https://www.axios.com/2026/04/07/anthropic-mythos-preview-cy...

- https://www.businessinsider.com/anthropic-mythos-latest-ai-m...

- https://www.reuters.com/world/anthropic-ceo-dario-amodei-arr...

Yes, this makes a lot of sense, but it’s just very amusing to see. 2 months ago, the world was about to end, now not so much.

Isn’t it just hilarious that a model that seemed so superior to Fable but didn't get doomsay marketing from Anthropic got released without any issues? In theory, this was supposed to be AGI level according to Anthropic, yet here we are, just a normal Friday.

+1. I've been only using Sonnet/Opus these days for UI work because GPT 5.5 just can't do any of that. Its just really terrible. Eager to give this one a try.

Thanks for the reply, Atty. So I guess I'm holding it wrong?

For some reason, I can't use the camera while in Live mode. The only option I see is the plus item, which does show the camera, but when I open it up and ask "Are you seeing my camera?" it will always say no and recommend me to open it.

Feels like the official camera icon does not show up for me? iPhone 13, ChatGPT Pro subscription.

Every time I get excited about Grok’s performance on benchmarks and demo videos, I test it myself and end up disappointed.

I'll give this one a try with a grain of salt and lowering my levels of expectations

I'm very eager to test this for brainstorming!

One thing I noticed is that we lost vision feature for some reason on the live chat?

This was an extremely useful feature. Not sure if it’s a regional thing or that they just removed that from the current live chat.

I imagine it will be even more useful with this new version.

Very cool. Not cool bringing Brazil’s loss to Norway again. We're already devastated. No need to keep beating someone on the ground. :(

What bothers me is that now I struggle to pass that "this is clearly AI slop" impression and just keep reading.

Sometimes the content written by AI is good, but it’s really hard to convince myself to just keep going when I detect it in the first 10 seconds.

"Consumers have developed pattern recognition for AI-generated content."

The irony of reading an article that talks about AI slop that clearly seems to have been written by AI. Hey, I could be completely wrong, and it wasn't, but there are so many flags.

Do I care? Not really, but whoever wrote this is right. I guess we developed a pattern recognition for these things

I can speak for myself. We are exactly at this moment trying to replace GPT 5 mini with an open weight / open source model. No luck so far.

Here is a trend I'm noticing:

- GPT-5 mini costs $0.25/$2 and will be discontinued in December.

- GPT-5.4 mini costs $0.75/$4.5 and is supposed to be the replacement.

- GPT-5.4 nano costs $0.2/$1.25 and, while it ranks better in benchmarks than GPT-5 mini, it's not even close when you test it in real scenarios.

So you're left being forced to go to GPT 5.4 mini if you use 5 mini today.

The same thing is happening here as their “Luna“ model will cost $1/$6.

Can't we just stay with the models we actually want? I don't need GPT 5.4 mini. GPT-5 does the job.

Maybe it’s the realization that it was never that cheap in the first place and they're forcing us to upgrade in a slow and painful way.

UX suggestion to make going thought this much faster:

1. Frame each option with one key (1,2,3,4). User press 2, select the second option

2. Let the user change options if they want until they press Enter. Enter submits the answer.

3. Once submitted, another Enter brings the next one

> What is a bit unique is: 1) we built it in Rust

The first unique characteristic is that it was built in Rust? Why does it matter from a user perspective? I was expecting the first point to be something that would convince me to check it out.

Unless the goal is to find people to collaborate on building the software. I got a bit confused.

Looking good regardless :)

That’s a supply and demand issue. If there were more clients than lawyers, I can totally see a reality where they speed things up because now in one week they can work on 20 cases instead of 2, thus 10x more money, but as you said, it doesn't seem to be the reality of the market.

edit: tldr; it does not seem in their best interests to be more efficient at this point

This is so cool! I'd love for it to have a front-page-like layout where "trending" news would have a bigger placement in the UI

Anyway, great work :)

> This article doesn't address writing code with AI, just code review. My issue with agentic coding is that I make numerous micro-architectural decisions while programming. I almost never have a full spec up front and develop one as I consider what I am writing.

working with AI forced me to write better specs but the way I write today is very different. I typically open Codex and have Linear MCP connected where my chat with the AI will end up writing the issue. Its a lot of back-end-forth where I tell what I want, the AI does all the code scanning, write something, I correct something, etc

The value for me is exactly that I tell what I want, the AI verify in the actual code if that's the path that makes more sense or not. In the end I have a pretty detailed spec that I'm much more confident is the correct path.

I find the spec easier to review than a huge PR so typically when executing is much faster and aligned with what I want.

The grill-me skill from Matt Pocock is great for this (https://github.com/mattpocock/skills/blob/main/skills/produc...)

I hate video calls, but sometimes is so much easier (and really faster) to align with someone over a complicated issue. feels almost counter intuitive but the more complicated the matter the easier to align over voice / video than text in my view

“Show me the incentive and I'll show you the outcome.” ― Charlie Munger

I’ve heard that using a VPN might cause in a ban. Considering they probably use an LLM to determine if you're a bot or something, you got unlucky with a possibly Haiku 4.5 hallucination ¯\_(ツ)_/¯

Anyone still use claude design? I’ve not seen any mentions on X, here or youtube recently, so wonder if it was all hype or people are actually using it.