RobIW

343 posts

RobIW banner
RobIW

RobIW

@RobIW_dev

15 yrs SQL + BI. Building the tools I want that don't exist(or they cost and arm and a leg). AIST protocol

Sweden 加入时间 Şubat 2026
29 关注15 粉丝
RobIW
RobIW@RobIW_dev·
“I’ve implemented a 10% slice and deferred the rest. Do you want to implement the rest?” Me: yes complete the rest “I’ve implemented a 20% slice and deferred the rest to future versions. This session is running long we should quit and pick up at a later time. Do you want to quit?”
English
3
0
9
322
Wyatt Walls
Wyatt Walls@lefthanddraft·
Opus 4.7, stopping 1 min into a task: Wait! Which Option would you like? A - This option is good and will solve your problem B - This option is bad and won't fix anything C - This option is just filling space I would recommend Option A. What do you want to do?
English
111
74
3.3K
122.6K
RobIW
RobIW@RobIW_dev·
@rez0__ I had 92GB of temp files from various debug and tests
English
0
0
1
1.1K
Joseph Thacker
Joseph Thacker@rez0__·
> my disk is getting full > i ask my agent why > The 82GB codex-tui.log 1) what
English
70
21
2.4K
260K
RobIW
RobIW@RobIW_dev·
@jturntdev It still one shots features for me but usage seems to have gone up. On the other hand my workflow has changed so I can’t confidently claim it’s not because of that
English
0
0
0
256
J J
J J@jturntdev·
Codex continues to degrade in real time. ChatGPT 5.5 is either been nerfed or under compute capacity constraints. Responses are taking forever, it feels like Claude Code all over again. That magic that 5.5 had, has gone. Just me?
J J tweet media
English
165
20
619
53.7K
RobIW
RobIW@RobIW_dev·
note to self: never let gpt 5.5 xhigh edit docs. it takes a fucking lifetime
English
0
0
0
8
RobIW
RobIW@RobIW_dev·
@fortelabs Have kids then you’ll never have to figure out how to stop being this kind of “poor”
English
0
0
5
561
Tiago Forte
Tiago Forte@fortelabs·
I think the main thing AI has taught me, through all the time savings it brings, is that I’m not a very interesting person Faced with a surplus of free time, I realize I don’t really have hobbies besides content consumption I’m forced to conclude that I don’t have very deep friendships, and am not a core member of any particular community I’m not very cultured, I’m finding, and don’t have abiding interests in art or literature or history or much that isn’t directly related to my work I have a work-centric life, in other words. AI pulls back the curtain on just how impoverished such an existence is, by disabusing me of its necessity Given the freedom I’ve always said I wanted, I’m at a loss as to what to do with it, except plow myself even harder into work, thus exacerbating the lesson There’s nothing more confronting to humans than freedom
English
387
233
4.2K
384.2K
RobIW
RobIW@RobIW_dev·
@onevcat Looks great but how do you hide away the sidebar and workspace columns? 30% of the screen is dead space
English
1
0
0
94
onevcat
onevcat@onevcat·
新版本的 Prowl，redesign 了部分内容，最主要的就是这个染色了...感觉每天用 Arc，整个人的设计理念都被“扭曲”成不同 workspace 需要不同颜色才舒服了..😂 prowl.cat
中文
13
13
261
38.4K
RobIW
RobIW@RobIW_dev·
Of course it makes sense Everyone is betting on their horse being the one that ends up like Microsoft excel. Used by every company at one point and most companies keep using it. You think launching an AI [something] like that for 200% addition on top is expensive? lol it’s cheap Every other such initiative has had much higher spend on top of what would ultimately be the real cost
English
0
0
0
38
Yoshik
Yoshik@AskYoshik·
The AI bubble math doesn't add up. Anthropic spends $3 to make $1 and that’s before you include any and all other costs like staff or electricity. Microsoft dumped $300B in capex, made ~$18B in AI revenue. OpenAI and Anthropic alone make up 43-54% of Microsoft, Google, Amazon and Oracle's entire revenue backlogs. Enterprises are burning through annual AI budgets in 4 months with zero measurable ROI. This is the most expensive science experiment in history, funded by your SaaS subscriptions.
Yoshik tweet media
English
370
996
4.8K
992.4K
RobIW
RobIW@RobIW_dev·
@elefarsee @cursor_ai Been using Claude only since opus 4.5 and switched to gpt 5.5 about three weeks ago and the difference was so stark you could tell already back then Claude was degrading hard
English
0
0
0
101
Luis Romero
Luis Romero@elefarsee·
Why is Claude Opus 4.7 turning into sh*t? It dodges many direct requests, it makes up a lot of stuff. Claude has been my favorite for 2 years now. I am now using Composer 2.5 heavily, not only because of my 'ties' with @cursor_ai, but because now I am dreading using Claude. @claudeai from a long-time fan, what's going on?
English
13
0
26
2.2K
user
user@user1508123·
@marcelkargul this has to be bait/ironic/sarcasm
English
1
0
32
3.8K
Marcel
Marcel@marcelkargul·
imagine if vscode looked this good
Marcel tweet media
English
71
8
510
57.9K
RobIW
RobIW@RobIW_dev·
@marcelkargul half the space is taken up by sidebars that aren't even filled and most of that information is dead weight. it's like having your bookmarks fill up half your browser window
English
0
0
3
1.9K
RobIW
RobIW@RobIW_dev·
@morew4rd I figured this would be your response. it's cool
English
1
0
0
7
moreward (moreboxed.com)
moreward (moreboxed.com)@morew4rd·
@RobIW_dev So you like to depend somone deciding to reset the limits for your tokens to have a reasonable cost? Lol and lmao
English
1
0
0
25
RobIW
RobIW@RobIW_dev·
@morganlinton @cjzafir gpt 5.5 xhigh is worse at planning than composer 2.5? I love 5.5 xhigh for planning but there is a better option?
English
0
0
0
280
Morgan
Morgan@morganlinton·
@cjzafir I still prefer Composer 2.5 for planning, Codex w/GPT 5.5 extra high for execution, that's the killer combo rn imo
English
10
1
30
4.6K
CJ Zafir
CJ Zafir@cjzafir·
Codex for planning Deepseek v4 for execution Trust me!
English
111
47
1.1K
61.5K
RobIW
RobIW@RobIW_dev·
@morew4rd Surely much cheap..Tibo: we reset limits Surely much ch…Tibo: limits been reset Surely muc…Tibo: limits are reset
English
1
0
0
28
moreward (moreboxed.com)
moreward (moreboxed.com)@morew4rd·
Kimi 2.6 + pi can work on a task for several hours on its own (if primed right) and fix issues. Zero extensions, raw pi + "medium thinking." Surely much cheaper than codex + /goal too.
English
9
1
71
5.1K
RobIW
RobIW@RobIW_dev·
@_dDeltaDt @BiggestComeback @BStulberg what is it that is hard to understand? steven says it lead to a chain of events that effected him in the next 3 days where he was performing sub-par compared to not having alchohol
English
0
0
1
20
Brad Stulberg
Brad Stulberg@BStulberg·
Y'all—this optimization stuff can make you fragile. If you are in recovery then yes, of course, a few glasses of wine will mess up your week (or worse). If you get drunk then yes, I could see it messing up a day or two. But if going out to dinner and having a few glasses of wine throws you for this much of a loop then perhaps you've actually just become fragile? I mean how would Steven manage having a newborn, or really any age kid? Or just the general uncertainty and messiness of life? In my new book I tell the story of golfer JJ Spaun, who was up all night with his vomiting toddler. His Whoop sleep score would have been zero. The next morning, he went out and won the U.S. Open. Actual excellence (not the elaborate, performative internet variety) demands resilience. It controls the controllables, no doubt. But it also ensures you don't optimize yourself into fragility, which is an increasingly common trap and performance killer. bit.ly/4uCzeQ7
Mikli@CryptoMikli

Steven Bartlett says a few glasses of wine ruined the next 3 days of his life “It's one of those areas where you don't understand the hidden cost until you really give it up for a while. I stopped drinking at 30 years old. I'm now 33. When I was 31, I thought, I'll have a drink again because now I could really A/B test it. I had a year of not drinking, decided to have a drink again” “It ruined three days of my life. I had a couple of glasses of wine, didn't get drunk. It ruined three days of my life because of the domino effect it caused” “I got worse sleep that night, and then because I got worse sleep that night, I ate more poorly the next day because my dopamine system or whatever, the cortisol system was all messed up. I podcasted worse. I didn't go to the gym that day or the day after because I felt really bad. I then slept worse, and I could track all of this on my Whoop”

English
117
130
3.1K
469.4K
RobIW
RobIW@RobIW_dev·
first results of the nightly improvement framework produced some great results.
RobIW tweet media
English
0
0
0
12
RobIW
RobIW@RobIW_dev·
new command palette is starting to take shape. man codex is so good
RobIW tweet mediaRobIW tweet media
English
0
0
0
16
Morgan
Morgan@morganlinton·
Okay, I'm so late to the party, it's time to just jump in and start - trying Pi tonight. Some people go out to dinner and drinks on Saturday night, some go to clubs, I try new agentic coding tools. I am probably the least cool person you know.
Morgan tweet media
English
41
3
168
23.7K
RobIW
RobIW@RobIW_dev·
@JohnShina91 @0xSero Works wonders right and an instruction to the implementer to check for the ref and present if they found something before implementing (ok this is more for auditing when something goes wrong anyways)
English
1
0
1
11
John Shina
John Shina@JohnShina91·
@RobIW_dev @0xSero I have a document titled Bug_Fix — where errors and the corresponding correct codes are logged for future reference.
English
1
0
0
15
0xSero
0xSero@0xSero·
You've been vibeslopping and every prompt fixes 1 thing and breaks 10 others It's okay, here's what you should do - In your IDE open the file viewer - Read the code, even if you don't understand it - Find dense repetitive logic and point at it - clean up, reduce lines of code
0xSero tweet media
English
14
3
118
5.5K

发现

@rez0__ @jturntdev @fortelabs @onevcat @elefarsee @cursor_ai @claudeai @user1508123