Register and share your invite link to earn from video plays and referrals.

Wes Winder
@weswinder
Building and 12+ years of software dev experience. Founder of the #1# vibe coding community on X. Toronto 🇨🇦
984 Following    12.8K Followers
of course claude is overloaded smh so frustrating when this happens they better give a reset since people are missing their 5h usage windows because of this
grok build is officially open source this is MASSIVE news grok build is easily the highest quality tui i've used they've def put a ton of thought and care into it feels like basically zero slop compared to claude code
Show more
We've open-sourced Grok Build and have reset usage limits for all users. Open sourcing Grok Build allows anyone to support making a reliable and robust harness. Check out our code, including the Git repo for the Grok Build CLI.
Show more
this timeline is much less addictive but x actually feels productive now
the for you timeline is so good now finally it’s people i actually follow
the for you timeline is so good now finally it’s people i actually follow
this is how you turn off data sharing with grok build simply use the "/privacy opt-out" command to turn data retention off it will also delete any previously synced data should 100% be off by default imo, but if you still want to use grok build, this is the best approach
Show more
forget tokenmaxxing tokenminning is the new wave i lowkey wanna see how far i can push the $20 codex plan + the $20 claude plan (+ free grok build with x premium)
i see this error way too often with 5.6 sol 😭
hot take: you should basically never use ultra effort with 5.6 sol in codex > it uses WAY too many tokens > your usage will get drained super fast (even on pro) > codex usage tracker doesn't update fast enough for you to react until it's too late learned this the hard way
Show more
there’s something deeply wrong with how codex is tracking usage always seems to be fine for a bit after a reset then slowly it regresses and your limits are gone in a few prompts something about the gpt-5.6 launch triggered this regression hope they find the deeper issue soon
Show more
ngl i’ve been having WAY more fun with grok 4.5 fast in cursor than gpt-5.6 in codex grok 4.5 is just so much faster and smart enough to trust the code it writes makes codex feel like molasses curious what 5.6 on cerebras feels like though (probably stupid expensive)
Show more
there’s something deeply wrong with how codex is tracking usage always seems to be fine for a bit after a reset then slowly it regresses and your limits are gone in a few prompts something about the gpt-5.6 launch triggered this regression hope they find the deeper issue soon
Show more
just blew through my gpt-5.6 usage in a few prompts didn't even realize until it was too late gg 😭 really need that hard reset from @thsottiaux soon
never bet against elon grok 4.5 is really good (and cheap) love to see openai/anthropic getting real competition
stop using loops start using queues trust me
0
191
801
20
Forward to community
this is what happens when the market finally realizes claude isn't gonna kill figma lol anybody who actually cares about design knows figma is the only real option
so are we getting gpt-5.6 on tuesday or thursday? who has the inside scoop
why is codex so obsessed with "isRecord"?
cmon openai drop 5.6 sol today you really want us to be stuck with fable all weekend? do it. you won’t.
anthropic should really just add a $500 plan > 100% weekly limits instead of 50% > fable stays available after july 7 > higher fable usage limits so many people would get this
i asked codex to analyze the new x algorithm and find the new meta strategy for posting on x this is what it said: Make posts that get fast, real engagement from the right niche, without creating negative feedback. The algorithm is trying to predict whether each viewer will like, reply, repost, click, dwell, watch, follow, or negatively react. So virality is not just "get likes." It is more like: "Can this post reliably make the right people stop, react, and pass it on?" Practical strategy: 1. Pick a clear lane The system uses topic signals. If you post randomly across politics, AI, memes, fitness, investing, and personal updates, the model has a harder time knowing who should see you. A focused account is easier to route into the right audience. 2. Optimize for replies and reposts, not just likes A viral post usually gives people a reason to say "true," argue, add an example, quote it, or send it to someone. Strong formats: "Most people think X, but actually Y." "Here is the mistake everyone makes with X." "I tested X so you do not have to." "The uncomfortable truth about X is…" "If you are doing X, stop doing Y." 3. Get dwell time The model cares about whether people pause. Use a strong first line, then enough substance that people actually read. Good posts often have: > a sharp hook > short lines > one clear idea > a payoff at the end 4. Avoid negative reactions The model also predicts "not interested," mute, block, report, and not-dwelled. Engagement bait can work short term, but if people quickly scroll past, mute, or mark uninterested, it hurts distribution. 5. Use media when it genuinely adds value The system has video/media signals. A chart, screenshot, demo, short clip, before/after, or visual proof can increase dwell and reposts. Random images probably do not help much. 6. Post into active moments Since the feed uses recent behavior and served-history memory, early traction matters. Post when your target audience is awake and active. The first audience teaches the system who else should see it. 7. Make the post easy to classify Do not be too vague. If the post is about AI agents, say AI agents. If it is about NBA, say NBA. Clear topical language helps routing. 8. Build a repeatable content shape Viral accounts often repeat formats: > strong opinion > useful teardown > simple framework > data-backed observation > story with lesson > visual proof > timely reaction inside a niche The current "meta" is probably high-signal niche posts with strong reply/repost hooks and low spam smell. Not generic motivational posting, not pure engagement bait, not random trend-chasing. The algorithm looks increasingly context-heavy: > who the user is > what topics they follow > what they recently engaged with > what they already saw > whether your post creates real positive behavior
Show more