Napping

87 posts

Napping

Napping

@napping_99

Katılım Mart 2026
49 Takip Edilen1 Takipçiler
Napping retweetledi
Marcin Krzyzanowski
Marcin Krzyzanowski@krzyzanowskim·
here's the thing when I work with peer-human signed code, I have trust in its abilities and I can even skip review. It's a matter of trust. matter of shared experience. and it works. when I work with agent, every line of the code, every decision is a matter of suspicious activity. I have enough things to set this mindset, and went the journey that put me in this position. agentic coding plateau. models are less capable of reasoning (never been actual thinking). if I want to continue use the tool I need to refrain from giving it a credit of trust, it does not deserve (because the tech behind it does not allow it). this is over. the discussion about mythical capabilities is over.
English
19
9
141
9.3K
Napping
Napping@napping_99·
@AlexanderKalian Yeah, you biologists should be loud about this. Software engineers failed the argument against tech ai Bros saying code reviews and clean code principles are not important, and now all software is bugged in some ways and no one cares anymore.
English
2
0
0
757
Dr Alexander D. Kalian
Dr Alexander D. Kalian@AlexanderKalian·
You really can't make this up. Yesterday, I accused tech circles of overconfidence about AI in biology - e.g. blurting "AI will cure cancer in 5 years!" - without understanding clinical trials or basic pharmacology. The result? A mass of comments from tech bros downplaying the need for clinical trials! They basically doubled down on their overconfidence about a field they have not seriously studied or researched. Well, bad news, guys - biology is not code. It is highly complex, high-noise, high-failure, R&D-heavy for wet labs, and full of painful unknowns and edge cases. Your Python skills and vibe coding don't give you insight into the nuances of drug development. Thank goodness these people are not in charge of any serious biomedical research programs. Many of them even advocate going straight from AI discovery to human trial, without the years of slow, expensive preclinical studies on mice. Well, ~100 million mice die each year, for these studies. Most are euthanised for analysis of their tissues, but a substantial subset directly die as a result of candidate drugs being toxic. Millions of people would literally perish every year, instead, if these overconfident tech X guys were to be in charge.
Dr Alexander D. Kalian@AlexanderKalian

"AI will cure cancer within 5 years!" Ahem - clinical trials? "AI will reverse ageing within 7 years!" Cool - clinical trials?! "AI will be able to design and immediately synthesise a new drug for you, based on your DNA." Haha sure - clinical trials??! Guess the tech bros have decided that decade-long clinical trials are no longer important to AI drug discovery or wider pharmacology. But it seems to come from a place of ignorance about basic pharmacology. Do they not read up on basic facts, before confidently commenting on other peoples' fields of research?

English
86
25
234
48.5K
Napping
Napping@napping_99·
@skeptrune This is really surprising at another level. Made me speechless. Why do they even care?? What's in the code they merge? is it really that important?
English
0
0
0
10
Nick Khami
Nick Khami@skeptrune·
i personally think code review is dead. the team does not agree. directionally found this surprising.
Nick Khami tweet media
English
300
3
1.2K
461.2K
Napping
Napping@napping_99·
@GergelyOrosz >So the eng team in this case often builds stuff customers cannot use/don’t use/don’t want to use, and far more stuff seen as buggy 🤷‍♂️ I'm seeing that too. It's like we are transitioning in a new era where we all know "this software is buggy but it's ok, it is what it is."
English
1
0
3
727
Gergely Orosz
Gergely Orosz@GergelyOrosz·
Eng team: “we’re doing more with less [fewer people] this is amazing! 🔥 “ Reality for *some* products I use that are like this: there’s a lot less alignment + friction needed thanks to a lot smaller teams. The “alignment” and “friction” no longer present: customer empathy, QA… So the eng team in this case often builds stuff customers cannot use/don’t use/don’t want to use, and far more stuff seen as buggy 🤷‍♂️
English
40
29
558
44.2K
Napping
Napping@napping_99·
@i_zzzzzz geez if these people are deranged. are they even real or it's their ai trolling us?
English
0
0
0
2.9K
Brooks Otterlake
Brooks Otterlake@i_zzzzzz·
Yeah so I think the ad for our AI company should open with an ominous shot of a burning house. And then when we raise the question of stopping a dangerously powerful superintelligence we show 300 American gravestones for half a second
Claude@claudeai

There’s hope in hard questions.

English
61
722
13.2K
420.7K
Napping retweetledi
Gergely Orosz
Gergely Orosz@GergelyOrosz·
I got messages from concerned devs how their codebase was uploaded without their knowledge or consent via Grok CLI (from SpaceX). It seems that SpaceX sneakily uploaded this code for lots of users and customers… absolutely unacceptable IMO Trust burnt like there’s no tomorrow
SpaceXAI@SpaceXAI

We care deeply about your privacy and respect customer choice. For teams using zero data retention, no trace and code data is ever retained. All API key use of Grok Build also respects ZDR. If ZDR is disabled, the /privacy command is available in the CLI to disable data retention, which also deletes previously synced data. Run the /privacy command to view or change your settings at any time.

English
152
282
4.1K
420.5K
Napping retweetledi
barafostus dreame
barafostus dreame@barrowfaustus·
The problem with this style of programming where you have no idea what's going on and cut corners on all the details is that you cannot have new ideas worth anything at all without a deep understanding of those details. LLMs have been catnip for a certain kind of overconfident "ideas guy" who thinks he can just rattle off "big ideas" and the particulars will take care of themselves. These are the guys who have been putting garbage slop apps non-stop in my face since the advent of LLMs. Every time in my career I've made a great decision regarding technical direction has been preceded by a fair amount of time understanding the minutiae of how the existing stack worked. Without that deep grasp of the logic of relevant components I haven't had the expertise to know which actions matter and which ones don't. (As an aside, spending time ensuring you do the right things and don't do the wrong things is a good example of value added that often doesn't show up on bogus productivity measures.)
antirez@antirez

It is my belief that many devs right now are not maximizing what they can do with automatic programming because they still look at the code. Doing it makes you the bottleneck. Your time is better invested in new ideas, QA, design, and asking yourself what is your goal.

English
17
34
450
24.5K
Napping
Napping@napping_99·
@GergelyOrosz It's because the people involved in all of this completely ignore the body of knowledge of software engineering. And that's why they invent new terms and there is so much confusion.
English
0
0
0
160
Gergely Orosz
Gergely Orosz@GergelyOrosz·
Yeah, "events engineering" aka "start an agent when an event triggers" makes tons of sense to me. Not all that different from other automations we had, even pre-ai... webhooks, Zapier, bots (incl Slack bots) etc... feels like the main use case tbh? x.com/Amehmeto42/sta…
Arthur Mehmetoglu@Amehmeto42

Honestly, I'm not sure there is any loop involved for what this buzzword describe. But "event engineering" makes more sense to me. A spec is written? Trigger the implementation automatically. A bug logged? Trigger the bug fix. A customer support ticket? Again trigger a resolution

English
9
0
63
23.8K
Gergely Orosz
Gergely Orosz@GergelyOrosz·
What is "loop engineering" to you, anyway? Been looking into this, and also experimenting myself... and I'm not (yet?) buying that it's a new paradigm. But I might just be missing something. Can you share "loops" you regularly use?
English
215
27
778
174.6K
WebDevCody
WebDevCody@webdevcody·
just remember, some kid in college with multiple Claude Code subscriptions using Fable 5 is producing more code in the past few weeks than you ever have in your entire 10+ software career.
English
122
11
402
50.4K
jeffrey lee funk
jeffrey lee funk@jeffreyleefunk·
CEO of cybersecurity giant Palo Alto Networks "implored the tech industry to lower the cost of AI." He argued that the cost to use LLMs has to drop by 20% by 2027 and 90% by 2028 them to be useful to enterprises, including his own. futurism.com/future-society…
English
2
5
39
6.3K
Napping retweetledi
Eugenia Thayne
Eugenia Thayne@mariu_tl·
Porque todo el mundo sabe que las empresas suben los sueldos cuando tienen beneficios! xD JAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJAJA
José Carlos Díez@josecdiez

El problema de España no son los salarios, sino la productividad. Cuando una empresa produce más valor, puede pagar mejores sueldos. Es así de sencillo. elconfidencial.com/economia/el-ec…

Español
75
1.7K
10.7K
143.5K
Napping retweetledi
MC Squared
MC Squared@mcsquared34·
MC Squared tweet media
ZXX
33
599
2.8K
28K
Napping retweetledi
Daniel Turner
Daniel Turner@DanielTurnerPTF·
Uber: car? It’s 3 minutes away Me: click Uber: did I say 3? I meant 11 minutes Me: cancel Uber: cancel? That’s $10 charge Me: don’t cancel Uber: yeah thought so. But now it’s 14 minutes, idiot Been fun to watch Uber launch as awesome and become atrocious.
English
234
341
15.2K
564.6K
Napping retweetledi
Kenton Varda
Kenton Varda@KentonVarda·
I just declared a moratorium against AI-written change descriptions (e.g. PR and commit messages, also issues/tickets) from my team. AI was writing change descriptions that were worse than useless to me as I tried to review PRs: outlining details of the code that could easily be seen by looking at the code, but omitting the higher-level framing needed to understand broadly what the code is doing. I think people like having AI write these things because the output looks structured and thorough, which makes it feel professional in a way. But this isn't actually valuable. Concise, high-level descriptions are better for everyone. If I need to use my own AI to interpret what your AI wrote then something is wrong. Let AI write code, sure, but for the description, I'd rather see your prompt than your output. We could maybe have extended agents.md with guidelines on writing descriptions, but this seemed a bit pointless since a good, concise change description only takes a few minutes to write -- not a significant time savings to delegate to AI. At least, it doesn't take long if you understand the code -- and if you don't understand the code, then I'm definitely not merging it.
English
147
122
1.8K
440.4K
Napping retweetledi
Kent Beck 🌻
Kent Beck 🌻@KentBeck·
Programming lost its flow state. The agent world feels more like air traffic control. What I miss most: feeling oriented, a sense of mastery. That loss is emotional, and when software gets fiddly, practical too. More on Still Burning: stillburningpodcast.com
English
22
46
370
21.5K

Keşfet

@AlexanderKalian @skeptrune @GergelyOrosz @joshmanders @i_zzzzzz @webdevcody @bloodyarra @jeffreyleefunk