Mike Rowave

Interested in Computer graphics, GenAI, Simulations and Gaming

[Mike Rowave @MikeRowave67]

@SolitudeRain_ Ai generated code is generative ai

The ai literally generates it. That's the definition of generative ai

You don't have to be such a performative hypocrite. It's fine to like something made by ai, you don't have to jump through hoops to justify it.

[Mike Rowave @MikeRowave67]

@Oklahoma_Brave DO NOT REDEEM!! DO NOT REDEEM!

[Mike Rowave @MikeRowave67]

@firesidealpha Ai bubble pop imminent?

[Mike Rowave @MikeRowave67]

@nivotino

[Mike Rowave @MikeRowave67]

@TrueSlazac Vr was cool too, but not the metaverse

[Mike Rowave @MikeRowave67]

@BernieSanders You're still a radical crazy extremist.

[Mike Rowave @MikeRowave67]

@bilal_akh

[Mike Rowave @MikeRowave67]

@sir_franco_ @Angaisb_ It's intentional. It's why ai detectors can detect GPT's watermarking but not something like krea 2, which doesn't have the same issues

[Mike Rowave @MikeRowave67]

@Polymarket As if he wasn't the #2 proponent of this message, second only to Dario

[Mike Rowave @MikeRowave67]

@qinzytech The tokenizer allegedly matches glm byte-for-byte

[Mike Rowave @MikeRowave67]

No office ever looked like this in the 2000s

[lost in latency @lostnlatency]

average 2000s office before minimalism took over

[Mike Rowave @MikeRowave67]

@bijanbowen @cgtwts Allegedly it was tested and confirmed to be 100% byte-for-byte accurate to GLM's tokenizer

[Mike Rowave @MikeRowave67]

@SharklettVT Can it run doom

[Mike Rowave @MikeRowave67]

Local LoRA training on AMD might be doable now, currently trying to train a krea 2 lora on 16gb

[Mike Rowave @MikeRowave67]

@aisearchio GLM-5.3 Air???

[Mike Rowave @MikeRowave67]

@EskilSteenberg Where'd all my bacon go :(

[Mike Rowave @MikeRowave67]

@thrillerhub I watched the entire thing and never even smiled

[Mike Rowave @MikeRowave67]

@trikcode Qwen 3.8 27B vibecoding go brrrrrrrr

[Mike Rowave @MikeRowave67]

@Trollacoaster And?

Are we supposed to care?

[Mike Rowave @MikeRowave67]

Qwen 3.8 27B is GPT-5.3 level

Not even exaggerating