๐ฎ Game Developer โ๏ธ 23 Creator of Day of Dusk, Oretary, Wizardium among other things. Comments are not reflective of any institutions I am associated with.
I might agree with what you post, I might disagree - but I enjoy seeing different perspectives which is why I follow & engage regardless.
Love the X community. โค๏ธ๐ซถ๐ป
Letโs all grow together! ๐
Opus 4.8 with Workflows is insane.
It can dispatch Opus 4.8 agents - there was 24 doing recall and another 64 doing reviews.
This is the ULTIMATE code base scourger.
Grok Build seriously needs a way to hold CTRL in terminal to fast delete and fast scrape text.
When you have a big prompt it's painful deleting the entire thing if you've changed your mind.
Grok Buildโs sub-agent spawning is wild. ๐คฉ
I just spun up 15 sub-agents, each with separate roles and responsibilities, and sent them across an entire codebase in parallel...
One mapping architecture. One hunting bugs. One tracing data flow. One checking tests. Others digging into security, UX, APIs, infra, and edge cases.
It feels less like โasking an AI to read codeโ and more like deploying a full engineering squad into the repo. ๐คฏ๐คฏ
Grok Buildโs sub-agent spawning is wild. ๐คฉ
I just spun up 15 sub-agents, each with separate roles and responsibilities, and sent them across an entire codebase in parallel...
One mapping architecture. One hunting bugs. One tracing data flow. One checking tests. Others digging into security, UX, APIs, infra, and edge cases.
It feels less like โasking an AI to read codeโ and more like deploying a full engineering squad into the repo. ๐คฏ๐คฏ
Grok Buildโs sub-agent spawning is wild. ๐คฉ
I just spun up 15 sub-agents, each with separate roles and responsibilities, and sent them across an entire codebase in parallel...
One mapping architecture. One hunting bugs. One tracing data flow. One checking tests. Others digging into security, UX, APIs, infra, and edge cases.
It feels less like โasking an AI to read codeโ and more like deploying a full engineering squad into the repo. ๐คฏ๐คฏ
Been working on something over at @jokelboard.
Basically it's a deeply integrated Kanban board into the Roblox ecosystem.
Will be shipping this soon to the DYLANJKL ECOSYSTEM and then potentially even customers!
Exciting times. ๐
@immasiddx Scary when you realise the valuations of these companies are within a balls throw of each-other and Anthropic still has 42x user growth potential over OpenAI - obviously this is Q3 2025 data; but still.
I know this might be a unpopular opinion with players but this system was used to exploit developers earnings by bad actors making their own template games to essentially get cashback on their Robux purchases (gamepasses, accessories etc) at the detriment of the developer/original item creator.
The only thing this will mainly effect is games that rely exclusively on commission income (basically low maintenance free cash flow games that rely on players donating to each-other with minimal creativity)
@bridgemindai https://t.co/U2C3dUJv1h
Would you still say OpenAI is providing better value here after benching it, or would you say you regret your decision to switch?
Just cancelled one of my two Claude Max plans.
Switching to ChatGPT Pro.
GPT 5.5 is the real deal.
And OpenAI's rate limits are so much higher it's not even funny.
$400/month on two Claude Max plans was never sustainable.
Especially when I'm hitting 100% session usage in under 2 hours on Claude Opus 4.7.
Additional feedback for Anthropic?
Release Mythos.
Still keeping one Claude Max plan.
But GPT 5.5 with Codex is earning more of my workflow every day.
@DavidBaszucki Hey, @DavidBaszucki
Does this support BYOM (Bring Your Own Model) yet and if not does Roblox have any plans to do so in the future?
This would be a deal breaker for me since I have custom Luau distilled models that outperform most mainstream models for Roblox development.