ashed96

5 ms·

1.
Show HN: LLM-Wiki but for Early Founders(github.com)
6 points by ashed96 5mo ago|2 comments
2.
by ashed96 5mo ago
GTM Engine turns Claude Code into your sales team. Eight opinionated slash-command skills that run your entire go-to-market motion — pipeline tracking, daily rituals, call analysis, lead scoring, cold outreach — all from your terminal.
3.
by ashed96 10mo ago
Glad to hear this, finally can use Claude for my projects! And missing sources: https://www.anthropic.com/news/claude-in-microsoft-foundry https://www.anthropic.com/news/microsoft-nvidia-anthropic-
4.
by ashed96 10mo ago
Junior makes random mistakes. LLM makes consistent mistakes. After a few weeks you know the patterns: where it's reliable, where it hallucinates.
5.
by ashed96 10mo ago
You're right - poor wording. Static vs dynamic is what I meant.
6.
Ask HN: Is Python's dynamic typing a liability when generating code with AI?
1 points by ashed96 10mo ago|3 comments
7.
by ashed96 10mo ago
Suddenly feeling better about our 99.9% uptime SLA. When even Cloudflare goes down, nobody can blame the little guys.
8.
by ashed96 10mo ago
No specific structure - you'll need to experiment based on your project. I treat it like briefing a senior engineer. Example: "Currently we have X, Y, Z but it has ABC limitation. We need to support FOOBAR. What are our options?&q
9.
by ashed96 10mo ago
From that standpoint, it's definitely happening already. But removal is easier too now. The real question is whether teams will have the discipline to delete. Most won't, but those who do will build better products.
10.
by ashed96 10mo ago
AI made us 10x more productive. Management noticed. Now we are expected to ship 10x more features. The free time was a lie we told ourselves.
11.
by ashed96 10mo ago
I'm building a startup myself and we vibe code everything, except core/complex parts where AI struggles. My workflow: 1. Give the agent detailed task + relevant file contexts (so it doesn't waste time searching) 2. Guide it w
12.
by ashed96 10mo ago
Once you've gone through the learning resources others shared, LeetCode has a dedicated concurrency section for hands-on practice: https://leetcode.com/problemset/concurrency/ (I actually authored a few of th
13.
by ashed96 10mo ago
It had been up for me, from Asia.
14.
by ashed96 10mo ago
Couldn't relate more. I built this Chrome extension using similar logic to youtube-dl: https://chromewebstore.google.com/detail/audio-only-youtube/... Eventually couldn't keep up with maintenance and lea
15.
by ashed96 10mo ago
My WHY: I need money to pay rent. My company's WHY: Generate shareholder value. There, saved everyone $97. You're welcome.
16.
by ashed96 10mo ago
Nice read though! Also checkout https://www.youtube.com/watch?v=iT7kq-R3Gjc too. It's absolute gold for GEO/AEO.
17.
by ashed96 10mo ago
What's your take on using AI-generated content for GEO? If it follows E-E-A-T and adds genuine value, does the origin matter for AI visibility?
18.
by ashed96 10mo ago
In theory, maybe to some extent yes - unique usernames could beat reused emails. But let's be real - nobody actually does that.
19.
by ashed96 10mo ago
All marketers ~~are liars~~ tell stories.
20.
by ashed96 10mo ago
Still on youtube-dl here - didn't know yt-dlp was the actively maintained fork.
21.
by ashed96 10mo ago
IIRC there's pre-processing (embedding/tokenization?) before feeding images to LLMs? Hit this issue optimizing LLM request times. Ending up lowering image resolution. Lost some accuracy but could bear that.
22.
by ashed96 10mo ago
What metrics are you tracking specifically?
23.
by ashed96 10mo ago
Opposite problem here - all building, no marketing. Now struggling with GTM because nobody knows we exist. The real trap is thinking it's either/or. You need both engines running.
24.
by ashed96 10mo ago
As others mentioned, you can't stop them from doing this. It's a cutthroat game where companies will try to poach customers however they can. Best defense is making your customers so happy they won't switch, but tactically: -
25.
by ashed96 10mo ago
IDM - Internet Download Manager -- Only available for Windows, never saw a proper alternative for Linux/MacOS with the same reliability and browser integration. Though no longer download videos/files enough to complain about it.
26.
by ashed96 10mo ago
Curious how you're thinking about the MLE role evolving with LLMs disrupting so much of traditional ML work? BTW, try Claude Code for your next project - you will love it! :)
27.
by ashed96 10mo ago
In my experience, LLMs tend to take noticeably longer to process images than text.