Bolumo

Recursive Systems Architect // Rational Visionary. Scaling intelligence recursively. Deconstructing LLM latent spaces to find the signal. 🛠️Building the future

[Bolumo @bolumo1]

@IterIntellectus nope the serious researchers arent going to use it to basically gift their knowledge to anthropic, to let them train their now models. it gets more and clear companies cant trust on them big ai labs

[Bolumo @bolumo1]

@rudraHQ here👏

[Sahil @sahill_og]

Me prompting Claude fable 5 at 3 am be like:

[ℏεsam @Hesamation]

Mythos, regulate something.

make no mistakes.

[Bolumo @bolumo1]

@kanavtwt if you use it right your mind will extend instead of decrease

[CooperBaggs 💰🍞 @edgaralandough]

I saw a girl say that when she feels stuck, she imagines she's 85 years old and gets one chance to come back to her younger self for one day, suddenly, everything feels like magic again, the walk, the music, the sunlight, live everyday like that.

wow. im doing this DAILY now.

[Bolumo @bolumo1]

@BenjaminDEKR distribution + Colossus utilization, walks on the acquisition once Q4/Q1 renewals print.

[Bolumo @bolumo1]

@BenjaminDEKR Almost nothing justifies $60B. The honest breakdown:

- VS Code fork → zero architectural moat
- No proprietary foundation model. Composer 2 is RL on top of Kimi K2.5, Composer 1 was built on DeepSeek (both undisclosed until reverse-engineered)

[Bolumo @bolumo1]

@BenjaminDEKR lagging metric — Q4 enterprise renewals will expose it
Real value = brand + ~1M DAU distribution + 12mo enterprise lock-ins. That’s a $10–15B asset.
The $60B-or-$10B option structure is Musk’s hedge. Smart money: he takes the $10B partnership for grok

[Bolumo @bolumo1]

@BenjaminDEKR Bundle problem: Claude Pro includes Claude Code, ChatGPT Plus includes Codex. Cursor is the $20 you stop paying first
•Leading indicators already cratered: Head of Eng left, Valon defection went viral, investor admits portfolio cos decoupling. ARR is the

[Bolumo @bolumo1]

@NuScienta @OpenAI they literally showcased the capabilities live on stream. prompted the model and showed the result.

what a time to be alive

[Bolumo @bolumo1]

@realBigBrainAI it is shifting, if we look for example at the new gemma models. theyre smaller and getting more effective for their seizes

[Bolumo @bolumo1]

what is meta doing? why arent there any groundbreaking announcements? zuck hired top talent for hundreds of million last year with no results??

am i missing something?

[Bolumo @bolumo1]

- steve jobs

[Bolumo @bolumo1]

you can fail at what you dont want, so you might aswell take a chance on doing what you love

[Bolumo @bolumo1]

keep building. just. dont. pause.

1 in a lifetime, lifetime.

[Bolumo @bolumo1]

when in doubt, take the way thats more uncomfortable - killing comfort

[Bolumo @bolumo1]

@OpenAIDevs have to keep reminding it to keep it short else it keeps on and keeps on yapping

[Bolumo @bolumo1]

@OpenAIDevs guys ... please feintune 5.4 for codex. great coding model but just keeps yapping in chat, its so hard to communicate efficient as with 5.3-codex. if thats fixed it would be an awesome model but rn just so annoying compared to 5.3..

[Bolumo @bolumo1]

@mazzenilsson @grok why are they losing 1m a minute