PhreeStyle

New account, so I guess that means I can make crazy predictions because I have no credibility to begin with 😂. This is a thread about #AI, $BTC, and $TSLA. What could possibly go wrong? 🧵👇😬

How is it possible for @github to have multiple outages per week now, and we're just cool with that now?

This is good. Old software analogy: shit in, shit out. This is a great thing for the quality of AI output going forward!

I have been a backend/platform dev a long time and my limiting factor has always been frontend. Just discovered v0 from Vercel. God help you all. 🫡

Why @opencode 😭?

@lucatac0 @thdxr Nope, but their benchmark shows usage under OpenClaw, and most importantly, the price.

@PhreeStyleBTC @thdxr GPt-5.4 is better at Opus-4.6 in social interaction? Are you sure about that

i haven't really clicked with the openclaw category of product so i'm having trouble understanding some stuff can someone help me understand why it needs a particular good model? isn't it doing more basic stuff?

Last week's civil war style reenactment of the Apollo era was kinda cool.

@melvynx Finally figured out why. Grok found this: rosspeoples.github.io/orpt-bench I've been looking for this exact thing for weeks! Cheap != inexpensive.

go usage go fast af

@chribjel @BniWael @opencode I love @opencode, but it does have some rough edges, and it's been a tight race with Codex and Claude Code (until a few days ago). The OpenCode Go service pricing seems a bit misleading at first, but also not entirely their fault.

@BniWael @opencode i am not looking for anything else than opencode. i just lay out about how i think opencode can improve their onboarding of new users

i am a dev who wants to use the best tools available, thus often using highly configurable tools. for these types of tools to gain mass adoption, they need to have really good defaults. IMO @opencode is not that. currently you get thrown in to a no-config setup that gives you access to "Big Pickle". this is a cool approach, letting people just install and go. but IMO they need an onboarding setup that encourages them to get the most out of OpenCode for the best possible experience. give them the easiest path to get access to Opus 4.6 and GPT-5.4

@chribjel @opencode I argue, it's free and therefore provides the best first-user experience versus "yay opensource, now pay up".

Y'all aren't gonna believe this. GPT-5.4-nano is the best value model for agentic workflows in @opencode. Scores the best, costs the least in my benchmarks.

I've been benchmarking the models. Yeah GLM-5.1 is decent, but it actually costs more to run agentic workflows than GPT-5.4. Why don't existing benchmarks surface this?

@melvynx Literally the next day, GLM-5.1 drops. This is the best timeline.

@melvynx GLM-5, MiMo V2 Pro. Recently found this helpful site: internlm.github.io/WildClawBench/

Search the Opus 4.6 replacement for my OpenClaw: day 1 I tried Kimi 2.5 today. It's not good. It has issues with tool calls. It outputs Chinese characters. It's bad... really bad. What's next?

I did not expect that. Solid move for Intel TBH.

@johnennis Same. Some people just wanna know when they can retire and have robots clean the house all day. Most don't think about it at all.

I think one of the biggest challenges when it comes to going hard into using AI is loneliness I am learning all these awesome things and becoming super capable But the set of people that I can really talk to about it is very small Is anyone else having this experience?

@weswinder How do you manage costs? You just accepting it?

@PhreeStyleBTC i have all the subs i don’t just use one

this is literally the best deal on the market rn for an ai coding plan > $10/month > kimi k2.5 > minimax m2.7 > glm 5 wayyyy more usage than $20 plans with big labs super underrated imo

We need real apples-to-apples “effective agentic tasks per dollar” benchmarks. Anything less is misleading at best. Who else is tired of this? #AICoding #OpenCode #LLM #AI #OpenClaw

Proof (official docs, April 6 2026): • GitHub Copilot (OpenCode): 1 premium request = one prompt YOU send. Autonomous agent loops & tool calls are mostly free. • ChatGPT Plus: Every user message/turn (including steering agents). • Routers: per-token or per internal step.

The AI industry’s “request” shell game is straight-up deceptive. Every provider uses the word “request” differently — yet every benchmark, pricing page, and “value” comparison treats them like identical units.