Hannes

Tinkering curious soul

[Hannes @za_hns]

@IvoVegter Ok, didn't know, thanks for that info.

[Hannes @za_hns]

@IvoVegter Please explain more, what conditions of employment were not met? You can't make a statement like that without addressing this obvious question.

[Hannes @za_hns]

Exactly, for example don't let the agent compute a hash function, it's a total waste of inference and that is not the type of computation that LLM's are made for.

[Uncle Bob Martin @unclebobmartin]

When using agents, ensure that everything that can be deterministic, is done with a deterministic tool. Don't try to get the poor agents to follow a deterministic process.

[Hannes @za_hns]

@kalliekriel @afriforum Oldest trick in the book, use semantics if you can't win an argument.

[Hannes @za_hns]

Can't wait for the 27b model!

[Qwen @Alibaba_Qwen]

📢Meet Qwen3.8-Max — our most capable model to date.

Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉

Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:

- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：https://t.co/iVHZWQoeSo
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.

💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens

Start building with Qwen3.8-Max! 🚀

📖 Blog: https://t.co/iwjmQxLBof
✅ Qwen Studio: https://t.co/4V2pFvDovG
⚡ API: https://t.co/gAGqaLQGbN

[Hannes @za_hns]

@chamath Framing it as a rewrite of all software is not practical nor necessary. Agentic dev code quality greatly depends on harness as well as processes around coding agents and code that is being produced autonomously. It forces a higher level of abstraction from dev teams

[Hannes @za_hns]

@lauriewired Interesting!

[Hannes @za_hns]

@GarethCliff 100%, well said, I hope we all have the courage to be principled

[Hannes @za_hns]

@WesRoth Coding can be thought as an abstraction of the world, meaning if you get good at the representation/simulation of something it will generally equate to improvements in real world applications AKA AGI

[Hannes @za_hns]

@WesRoth Consider ontologies of decision-making, I think you will find the role of "evolutionary betterment" a fallacy.

[Hannes @za_hns]

@sickdotdev Wow you must be living under a rock, no offence

[Hannes @za_hns]

I have experimented with a lot of open source models that I can actually host, but none of the "mutated" qwen 3.6 27b variants come close. I don't know what Qwen did, but the 3.6 27b is a beast on coding tasks.

[Hannes @za_hns]

Decided to give X a go! Does the recent influx of influential people on the platform have anything to do with the release of Grok 4.5?