Modes

[Modes @modesioo]

@MiclauMarius @elonmusk Mars

[Modes @modesioo]

@kimmonismus I wonder if it will be open source

[Modes @modesioo]

@bridgemindai I think this is Anthropic’s problem. They’re losing money on that model at the current subscription price.

[Modes @modesioo]

@Teknium Since you said “good luck next time”…when’s the next one?

[Modes @modesioo]

@teslaownersSV @elonmusk I dont think it would be second if GPT 5.6 were listed there.

[Modes @modesioo]

@orcdev @shadcn Curious, did you use an MCP connector for Unity, or was the agent only modifying the C# project without directly controlling the Unity Editor?

[Modes @modesioo]

@bridgebench @bridgemindai I think it will be something between Fable 5 and GPT 5.6, but cheaper.

[Modes @modesioo]

When I was 20, I set myself a goal to make my first €1M by the age of 30.

Today is my 25th birthday, which means I’m exactly halfway there.

During those five years, I finished my master’s degree and started working as an AI developer. I currently make around €3,000 a month, which is a pretty good salary in my country.

But I’ve realized that a good salary mostly makes life more comfortable. It doesn’t change your life drastically, and you can still feel stuck in the rat race.

At my current pace, I’m still very far from the goal I set five years ago.

So now I’m trying to decide what to do next.
Should I keep my stable job, continue growing my career, and build projects on the side?

Or should I take a bigger risk while I’m still young and don’t have many responsibilities, lock in, and fully focus on building products or a business? It feels like there’s never been a better time to build, so part of me thinks it’s now or never.

I’d really like to hear from people who have been in a similar situation.

What did you choose? What worked for you? And what would you do in my situation?

[Modes @modesioo]

@byte_crypto_ @bridgemindai he's probably burning them more money in tokens than all 3 subscriptions are worth

[Modes @modesioo]

@bridgemindai Do we have also SWE-bench comparision GPT 5.6 vs Fable 5 ?

[Modes @modesioo]

@bridgemindai wow, that's rough. Planned to upgrade to Max today just to get some last use of Fable, but after seeing this I honestly don't know if it's worth anymore.

[Modes @modesioo]

How important is proper model-specific prompt template formatting in llama-index RAG solutions?

I'm working with model-specific tags (like Gemma's <start_of_turn> or Mistral's [INST]).

What are best practices for custom prompt templates with retrievers? Any performance gains?

[Modes @modesioo]

How many RTX 4090s are needed to match the LLM inference performance of a single A100 or H100?

[Modes @modesioo]

@AnandButani True, it’s tricky. Are there any guidelines or tips to make it less of a balancing act?

[Modes @modesioo]

What’s the best way to estimate on-prem hardware (GPU, RAM, etc.) for running an LLM like Mistral Small 3.1 (24B parameters) at decent speed for 100 active chat users at once?

[Modes @modesioo]

@MistralAI Better than specialized models like Tesseract or PaddleOCR ?

[Modes @modesioo]

@Gypsyforfreedom 🤣🤣🤣

[Modes @modesioo]

@bdlowery @RhysSullivan it looks normal for me ig

[Modes @modesioo]

@fast_to_market Toptal