Tom Arnold

@plainionist Nope - I don’t miss sifting through stack overflow at all lol

Could you imagine going back to coding without AI? 🤔

@iatnon @BillyBowss @ClaudeDevs Unfortunately it's the correct answer though. I lasted about 2 days on the 5x plan last summer before needing to bump up to 20x.

@BillyBowss @ClaudeDevs It’s so expensive…

I'm actually devistated... What am I supposed to do the rest of the week?!?! Only half way through this week and ran out of weekly limits. CLAUDE WHAT HAVE YOU DONE!?! Never even came close to weekly limits before @ClaudeDevs please fix this!

@Bhavani_00007 The Claude Vs. Codex mindset is wrong - you should be using both to their strengths.

I'm a Claude user. Give me one reason to switch to Codex

@davepl1968 The 1500 likes on that post is concerning 😑

If lines of code were a meaningful measure of software quality, then the ideal program would be an infinitely long, labyrinthine monument to redundancy in which every variable enjoyed its own paragraph, every conditional branch was lovingly expanded into a dissertation, every reusable function was duplicated dozens of times for the sake of numerical abundance, and every simple idea was buried beneath geological strata of boilerplate, thereby rewarding verbosity over clarity, complexity over elegance, maintenance burden over maintainability, bug surface area over correctness, and programmer ego over engineering discipline, while simultaneously ignoring the inconvenient reality that the history of computing is littered with brilliant achievements—from hand-crafted assembly routines to elegant algorithms and compact operating system components—that derived much of their value precisely from accomplishing more with less, making lines of code roughly as sensible a productivity metric as measuring the quality of a novel by its weight, the efficiency of an engine by the amount of metal in the block, or the intelligence of an argument by the number of words it takes to arrive at the point. Brevity is wit.

@AnthropicAI Would be interested to hear the details on what it took for the quality of code produced to to be on par with human code. Like what is the baseline for labor hour input for AI produced code to match good human produced code for the same objective.

The speedup isn’t just in volume. On open-ended coding problems where answers are unclear, Claude’s success rate is now 76%—a 50 point jump in just 6 months. Many engineers also say Claude’s code quality is now on par with human code; we expect it to be better within the year.

Our internal data shows Claude is accelerating AI development—a possible path to recursive self-improvement, or AI autonomously building a more capable successor. It’s happening faster than we thought, and the implications deserve greater attention. anthropic.com/institute/recu…

One of the more annoying issues I surfaced earlier this week for some of my long running Claude Code workflows is validator/verification agents focusing almost entirely on issues present within specs/code/reports etc. and overlooking what's not present, like issues you'd expect uncovered during a gap analysis. So Monday and Tuesday of this week I was running into the same types of issues you've outlined: Architect or another agent finds bug or issue > creates work order > work order processed by validation agent > work order updated by validator > implementation agent implements > something overlooked and new bug implemented > cycle repeats until complete. The goal of implementing the fix is ultimately achieved but required additional loops = opportunity for efficiency improvements. The fixes have been simple enough for Claude in this case - I needed to improve my prompting/instructions to consider not just the code or artifact being validated, but to also think about what is missing, or outside of the immediate blast radius. That might sound obvious but things were working well previously and I liked the way my workflows were running for several weeks, with minor adjustments here and there. Not sure if this is more a quirk specific to 4.8, but I can see this got me multiple times in a couple of days for my Claude Code workflows after the release of 4.8. Of course, any new model release requires some adjustments to skills built on older releases. I just found your post interesting with what I've observed recently, as well. 🙂

My AI agents are starting to feel like a real team. Architect identified a potential bug and asked Verifier to confirm. Verifier confirmed. Implementer fixed it. But the fix broke another scenario. Verifier escalated back to Architect. Architect reviewed and approved the next step. Eventually everything passed. AI teamwork 😉

@plainionist I love hearing interesting ways ppl implement in inter-agent collaboration. What is your framework for this team?

@ClaudeDevs Thank God.

We've changed the trigger word from "workflow" to "ultracode". You can still say "use a workflow for this", but when you're clearly referring to something else, Claude won't kick off a dynamic workflow. For an explicit trigger, use "ultracode". We appreciate the feedback!

@Occam1111 @ThePrimeagen Oh you must be the life of the party in your sad little social circle

@FoundryLabsTom @ThePrimeagen I could not get into it. Just feels so teenage edge lord

Looking for new fantasy book: * I loved Wheel of time * Liked mistborn * Loved storm light archives book 1/2, liked 3/4, meh on 5. (Just turned Kalladin into a soyboy) *Another kingdom enjoyed What else should I read?

@BachelderDan @ThePrimeagen Such an amazing series.

@ThePrimeagen Dungeon crawler carl. The best audiobook series I've ever listened to.

@wbuxtonofficial He still has time! Who knows, maybe his Indy 500 experience this year will inspire him.

What I’d have given for seeing Ricciardo run the 500. Think he’d have absolutely loved it out here. In another lifetime maybe. We can dream.

I've been enjoying Opus 4.8 more than 4.7, but there are still some annoying issues and performance quirks, partly related to the Claude Code harness. But overall, it feels better... Still hesitant to use the API in my own harness due to cost though.

@theo There's no incentive for them to improve efficiency while enterprise corps are still willing to torch absurd amounts of money on tokens... For the same results they could achieve with their own harnesses and decent model routing, at a fraction of the cost.

Resubbed at the $100 tier to try 4.8 and the new "ultracode" feature. I hit my limits in a single prompt.

@theo What kills me is how they've rebranded Skills (put your instructions in a directory) as "Workflows!" My new favorite command stub: /Skills-on-LSD > workflows

wtf this is real?? they really are leaning into the whole slot machine thing huh

Most workflows(skills) I build are essentially a large task list performed by multiple different models. When in CC it's sonnet/opus. Each task in the flow requires the agents to output their results in its own file in a directory. This happens for every task. Every workflow run can produces between 6-40 artifacts that can be analyzed later by another workflow. I execute anywhere from 20-100 workflows a day. This produces a large dataset I can audit and analyze. I built a hallucination check skill called `/badtrip`. It runs sonnet/opus against each other to catch hallucinations. Not only do I catch hallucinations, but I produce scorecards on model performance when validators aggregate results. I asked CC to visualize this skill and got this mediocre wall of markdown I can't even share in one screenshot.

@FoundryLabsTom @haider1 How do log and audit?

wait, is it just me, or opus 4.8 is getting dumber?

@davepl1968 Hell yea, good for you.

Dude, I retired like ten years ago. Since then, I've written and published two books, a metric crapton of software, learned 3D graphics and embedded systems, built a YouTube channel with a million subscribers, restored two cars, sent three kids off to college, learned to drive a race car, and now I'm a public speaker for fun. I don't have a garden. No chill needed.

@theo So far, from what I've experienced, the only thing you're missing is a buggy CC experience. Half of my long running workflows had API errors over the last hour, and it still outputs walls of nonsense text like it just had its first toke and can't shut up.

Good that Anthropic isn’t dropping anything right now, the day my Claude code sub expired…

@claudeai Cool, now if you could fix this bug I've been running into for the past 2 hours that would be great.

Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors. Available today at the same price.