frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Ask HN: Am I missing something with AI

3•vasko•1h ago
I constantly hear developers around me talk about how AI has completely changed their life and how they don't even program anymore, they just prompt. But any time I've used it, the output has always been off. And when the output is off I have to go and read through everything, learn how it works and fix it, which at that point I might as well write it myself.

I just don't understand what other people are seeing, I've mainly used Claude and ChatGPT, I got a free trial for premium but it's just underwhelming, their only use so far for me has been as a search engine, but they're a search engine that's wrong 20% of the time so even that use is questionable.

Comments

dejan_kocic•1h ago
I think AI is good for creating a foundation, then branching out and adding features, you shouldn't overdo it with AI.
morisil•1h ago
I am shocked how much my experience is different from yours. I wrote Claudine, my own version of Claude Code, almost 2 years ago. This experience gave me the understanding of how the technology works. Since then I've produced maybe 300k lines of open source code, and all of it meaningful to the bones. What kind of projects are you working on, maybe it's the specificity of your domain?
JohnFen•1h ago
You might be even more shocked to learn that the author's experience isn't rare.
Festro•1h ago
We're reaching a point currently where output quality is very much determined by input quality. Previously output quality was hampered fundamentally by model knowledge, hallucinations, and model quality.

Now, we have better knowledge of prompting as people have learnt what to say, models are better, models make use of memory from other conversations, they have skills written by humans or even themselves on how to do things, access to the internet to get live info, access to project files to check info, and the built in 'thinking' to challenge their own assumptions and loop on outputs until its refined.

You're right that output is always off still, but a lot of people have reached a point where it's only 'off' by an amount that is less than the effort required to do the task themselves, and considerably so.

My example today is prompting Claude to do a technical audit of a new client site.

It has skills for UX and SEO audits. Connects to an SEO tool. Pulls client info from OneDrive. Outputs to Word from a template for our agency. I even had it drive a remote pagespeed testing tool in Chrome because they don't have an MCP server currently.

Doing that report myself is 3.5-7 hours depending on what's found. Claude did it in 0.5 hours. Now I'm sorting out the oddities and anything that feels 'off'. I know and understand the full content of the report and can get on with actioning the recommendations or prioritising them for others. I've got maybe 1 hour of review and writing to do. It's not a 10x improvement but I'm happy with it.

Although, whilst Claude did it's bit I was doing other work. So, perhaps the multiplier is higher than I give it credit for.

jr_isidore•59m ago
When you join a new company, is it faster to fix a bug rewriting everything from scratch or to modify what's there? Seriously, get your head out of your ass.

Turbocharged Earnings Are Pushing Stocks Higher. There's a Catch

https://www.wsj.com/finance/turbocharged-earnings-are-pushing-stocks-higher-theres-a-catch-100839c9
1•matltc•54s ago•0 comments

Assembling the CISD (Colin's Integrated Serial Device)

https://www.colino.net/wordpress/archives/2026/06/19/assembling-the-cisd-colins-integrated-serial...
1•ibobev•1m ago•0 comments

Fixing a Quicktake 150 serial connector

https://www.colino.net/wordpress/archives/2026/06/21/fixing-a-quicktake-150-serial-connector/
1•ibobev•1m ago•0 comments

Lobachevsky's Integral Formula

https://www.johndcook.com/blog/2026/06/22/lobachevskys-integral-formula/
1•ibobev•2m ago•0 comments

Code Review by Dines and Gilfoyle

https://github.com/v1r3n/dinesh-gilfoyle
1•opiniateddev•2m ago•0 comments

Eli Lilly gave extraordinary obesity drug access to a 79y/o patient. Who was it?

https://www.statnews.com/2026/06/23/eli-lilly-unusual-weight-loss-drug-trial-compassionate-use-re...
1•johnhamlin•3m ago•0 comments

The $400M machine powering the future of chipmaking

https://www.technologyreview.com/2026/06/23/1138837/asml-400-million-dollar-machine-powering-futu...
1•Brajeshwar•3m ago•0 comments

When falling housing prices are good news – and when they're not

https://text.npr.org/g-s1-129339
1•mooreds•4m ago•0 comments

Europe's housing shortages are worse than America's

https://www.worksinprogress.news/p/europes-housing-shortages-are-even
1•speckx•6m ago•0 comments

Reimplementation of Mario Kart Wii's physics engine in C++

https://github.com/vabold/Kinoko
3•klaussilveira•6m ago•0 comments

Show HN: SwissReach – Browser-based isochrone maps for Swiss public transport

https://github.com/filippofinke/swissreach
1•filippofinke•7m ago•0 comments

Show HN: The Cascade Graph – An interactive map of AI and energy constraints

https://atomprophet.io/tools/cascade/
1•antisyzygy•7m ago•0 comments

Show HN: Agent skills that review user-facing agent UX from your codebase

https://github.com/Correl8AI/skills
1•romz•8m ago•0 comments

Likely Math Behind Subquadratic

https://github.com/jonsmirl/ssa
1•jonsmirl•9m ago•0 comments

We started gathering product feedback from visiting agents

https://www.sanity.io/blog/how-to-get-product-feedback-from-agents
1•jarodreyes•10m ago•0 comments

Generate vehicle bill of sale documents online

https://www.billandsale.com
1•xiyan•10m ago•1 comments

Sunsetting a Package Manager

https://nesbitt.io/2026/06/23/sunsetting-a-package-manager.html
1•chmaynard•10m ago•0 comments

Why marketing AI hallucinates: how can we ground it in platform-native data

https://www.fuse.is/blog/how-fuse-talks-to-tiktok-ads
1•rkovashikawa•11m ago•0 comments

Two Indexed Hash Tables

https://vnmakarov.github.io/data%20structures/c/c++/open-source/2026/06/23/two-indexed-hash-table...
1•theanonymousone•11m ago•0 comments

Blaiso Launches Today

https://www.blaiso.com/
1•msass•14m ago•1 comments

Why American data centers can't plug in

https://worksinprogress.co/issue/why-american-data-centers-cant-plug-in/
3•surprisetalk•14m ago•1 comments

Squadron 42 Release Announcement in October?

https://nosygamer.blogspot.com/2026/06/squadron-42-release-announcement-in.html
1•speckx•16m ago•0 comments

Takes more than an hour to deploy code to test?

1•GamingAtWork•16m ago•0 comments

Sparky – pocket size CI server and workflow manager

https://github.com/melezhik/sparky
1•melezhik•17m ago•0 comments

Stairwell In C# with Ultracontrapipe in A (Lydian) [video]

https://www.youtube.com/watch?v=23G5QDWqDUY
1•starkparker•17m ago•0 comments

Show HN: Project Cherub – A Forked TempleOS. Early Build and Future Plans

1•Rubinoslaw•17m ago•2 comments

Show HN: Publish.my – Static hosting where the AI agent is the customer

https://publish.my/
1•aizuikmal•17m ago•0 comments

AI's Affordability Crisis

https://blog.dshr.org/2026/06/ais-affordability-crisis.html
10•ilreb•18m ago•1 comments

Ask HN: How often do you use "Answer now" when using LLMs?

1•dwa3592•18m ago•2 comments

Show HN: An open MCP that gives coding agents page-cited embedded datasheets

https://github.com/ByteAsk/ByteAsk-Embedded-MCP
1•anirudhak47•19m ago•0 comments