justiceforsaas

7 ms·

1.
GLM-5.2's Code Reviews Are Only as Good as Your Prompt(blog.kilo.ai)
6 points by justiceforsaas 3mo ago|0 comments
2.
Claude Fable 5 vs. GPT-5.5: better planning, similar execution(blog.kilo.ai)
4 points by justiceforsaas 4mo ago|0 comments
3.
by justiceforsaas 4mo ago
Do you have an in-between step when you go from planning to execution? Eg. have a PLAN.md file you then ask the execution model to implement.
4.
by justiceforsaas 4mo ago
Honestly the code gen part has been “good enough” for a while now, especially with models like Opus. The broader point this post is making is that newer SOTA models are improving at the "planning layer", and this is usually the th
5.
Claude Fable 5 vs. GPT-5.5: Better Planning, Similar Execution(blog.kilo.ai)
20 points by justiceforsaas 4mo ago|9 comments
6.
SOTA Models Should Learn to KISS (Claude Fable 5 Experiment)(blog.kilo.ai)
1 points by justiceforsaas 4mo ago|0 comments
7.
Agents.md may trick us into writing better docs(blog.kilocode.ai)
28 points by justiceforsaas 1y ago|5 comments
8.
by justiceforsaas 2y ago
Are you suggesting these AI agents can write code 24/7 without a single error or getting stuck? In my experience (and I think plenty of others would confirm this), you're lucky if they can run smoothly for even 5 minutes without h
9.
No OpenAI, I Won't Pay $10k/Month for Your AI Agent(thetrendysoftwareideas.substack.com)
13 points by justiceforsaas 2y ago|3 comments
10.
DuckDuckGo to start providing "AI-assisted" answers(theverge.com)
6 points by justiceforsaas 2y ago|1 comments
11.
by justiceforsaas 6y ago
That's also what I thought. Instead of using fire extinguishers, FB/Apple/Google are adding more gasoline to the fire. This is going to radicalize people (especially conspiracy theorists) who now think have actual, indisputab
12.
by justiceforsaas 6y ago
Facebook is probably a bad acquisition channel for many SaaS businesses, especially after this week news on FB managers trashing their own ad targeting [1]. I do a little project where I analyze founder interviews, trying to discover which
13.
by justiceforsaas 6y ago
> We opened the floodgates. We shared the project on Twitter and Instagram, posted to HN and wrote a medium post. But, try as we might, no one was signing up. Later in the post: > Had we had more funding, we would have taken the time
14.
Acquisition Channel Opportunities #001(zerotousers.substack.com)
1 points by justiceforsaas 6y ago|0 comments
15.
by justiceforsaas 6y ago
Guys, add a search box where people can enter their site & see "if they quality". If they don't, and enter their site again (and are <1MB), maybe you can add them in a "Recent improvers" list. I think this wi
16.
by justiceforsaas 6y ago
I've looked at the themes and it looks like many of them were inspired from MRA (Male Right Activists) community. I think it's an important day to raise awareness of some important issues that men face in today's society, inc
17.
by justiceforsaas 6y ago
Just curious: What's the difference between Substack and RSS? With RSS, you 'subscribe' -> get blog posts to your news reader. With Substack, you 'subscribe' => get blog posts to your inbox. Of course, they al
18.
by justiceforsaas 6y ago
Is there a particular logical reason why Facebook required these people to get back to office? Is it related to productivity? Would be curious to hear Facebook's take on this.
19.
by justiceforsaas 6y ago
In one comment you said that the contrarian dynamic works on HN threads and to internet comments in general. I just went and opened 10 random Reddit threads on the front page to read the top comments, and found only 2 where this was true, a
20.
by justiceforsaas 6y ago
Oh, sorry, I haven't seen it. Saw it now. Is there a way to get in touch?
21.
by justiceforsaas 6y ago
The gist of the article is: The OP created a tiny project, launched it on Product Hunt/Hacker News, and moved to creating another project. I'm glad he acknowledges the problem with this: "A Product Hunt launch can easily get
22.
by justiceforsaas 6y ago
> I’d bet there’s few “micro-bets” that are generating any meaningful amount of money. That's the whole point of micro-bets. You acknowledge the fact that a small % of them will generate any meaningful $, so you deliberately minimiz
23.
by justiceforsaas 6y ago
I think you've gone from one extreme to the other. Yes, building a product with no distribution sucks. So does doing distribution without having anything to show for it. Take this from a guy who spent 2 years studying distribution chan
24.
by justiceforsaas 6y ago
Literally describes 100% of IndieHackers threads about "I sold 540 copies of my products on Twitter in 8 hours".
25.
by justiceforsaas 6y ago
Why is everyone so obsessed with ideas? Out of everything needed to succeed with a startup (idea, audience, acquisition channels, competition), the "startup idea" is the thing you can ALWAYS change. It's the thing you have th
26.
by justiceforsaas 6y ago
From reading the article, the main reason seems like having (way) less competition. Especially with cheaper acquisition channels (I've been doing research [1] on this topic for 2+ years and noticed this trend), where you have less busi
27.
by justiceforsaas 6y ago
It's painfully obvious this article was written by an investor for several reasons: 1) They define "growth" as a stage that goes AFTER raising money. As a founder, growth (planning, executing) should come before you ask for a
28.
by justiceforsaas 6y ago
Yes. Their app store is also a pretty good acquisition channel (I do research on this topic [1]). Here are some relevant quotes I found from my DB: 1. Credible ($2.5k/mo), a Shopify app to display recent sales: "The other primary
29.
by justiceforsaas 6y ago
This is an interesting case study, but my research on acquisition channels that consistently work for founders [1] has shown that sponsoring blogs/web-based "real estate" is hit-or-miss (and doesn't work) for most founde
30.
15 Acquisition Channels I Discovered from Reading 479 Founder Interviews(zerotousers.substack.com)
7 points by justiceforsaas 6y ago|0 comments

More ›