hite.sh

feedback is a gift

"Compacting conversation… " is the "Updating windows.." of the new world.

OpenClaw needs RLM more than it needs LLM. to keep the cost low.

@paraschopra I think its more to do with the initial launch where it came as POC of what all can AI actually do if application layer is planned nicely. I think the hype was helpful in shaping market around application layer from LLM. Claude just proved it.

I don’t get the OpenClaw hype Connecting Claude with Telegram / WhatsApp is trivially easy, you can literally ask it to help you do this and it’ll guide you. Same story with recurring jobs. I just did this - now Claude send me local bangalore news summary at 12pm IST daily on Telegram. Took me 15 mins to build. If the argument is that Claw lets nontech users do this, imagine the security implications when users let an LLM take over their system while having no idea what’s happening under the hood. Making custom scripts and workflows with Claude lets you at least know what you’re configuring on your system.

if something doesn't make an LLM deterministic its not a skill. Imagine a kickboxing hook/jab/cross that lands only 30% of the times.. its luck not skill

can we stop shipping .md files and calling them "skills"? it's anything but a skill. if an LLM can write its own .md in five minutes, that's not skill/learning/improvement, not anything. this isn't what self-learning was supposed to mean.

@anthonyazrak hey we are building banking for AI Agents. would love to learn from your experience on speedrun

speedrun changed my life + gave me some +10k cheques to throw at yall hit me up if you’re interested

@kenanhsaleh @a16z @speedrun hey, let's talk! we are building in AI Finance

I'm opening my calendar for 1:1 office hours today and next Friday Applications for the next @a16z @speedrun cohort open next week - come AMA before they do. Happy to chat about your idea, fundraising, or application strategy Reply and I'll DM a Calendly link for 15-min slots

did @GroqInc just compared state of the art 1T P model with a shitty 120b P model?

because of github.com/hiteshjoshi/la… which uses firepass as act as recursive language model. again, That's $14 of daily tokens just being used as mere thinking..

its been one week of using firepass from @FireworksAI_HQ and I use 44m tokens daily. That's $14 of daily tokens just being used as mere thinking.. (at a subsidized cost of $7/week) Final results are incredible.

something tells me.. leak of Claude code will be a turning point in the VC. Software is pretty much unsolved atm.

@euboid @Railway damn! I read blog.railway.com/p/incident-rep… I think it was railway. WTF!!

@euboid @Railway this seems like a architectural nightmare and not infra providers fault. Why would railway be responsible for this?

I love @Railway. Been a happy customer for 5 years. But today a caching misconfiguration let users see and act as other users. Names, personal data, medical info leaked between customers. This after weeks of almost daily incidents. It's time to switch. Alternatives?

#claude-code-cli-optional" target="_blank" rel="nofollow noopener">github.com/hiteshjoshi/la…

fireworks dropped Fire Pass, inference on kimi k2.5 is stupid good so i merged RLM-style long thinking with claude code + fire pass now i can actually walk away from my screen and come back to code that isn't slop

they are not even skills. merely "roles"

The more plausible code you have the more shit is coming! There is a reason we call it tech debt.

@garrytan G, LLMs are designed to generate tokens. LOC for you == just tokens for LLMs. It's purely fictional to assume LOC means good output. LLMs write plausible code not efficient code. The more plausible code you have the more shit is coming! There is a reason we call it tech debt.

Absolutely insane week for agentic engineering 37K LOC per day across 5 projects Still speeding up

@aidenybai Nothing to cope. Just acceptance. Death is the ultimate truth. It's actually the only truth.

idk how to deal with this so here goes nothing i have extreme fear of death the thought of ceasing to exist forever fills me with a terror beyond comprehension i also don't believe in god / afterlife how does one cope with this

literally anyone can ship quickly if they sacrifice reliability. it’s not in any way impressive.