maximedupre

7 ms·

31.
by maximedupre 1y ago
Are you saying LLMs are broken? This is nothing more than a wrapper. Not sure why it triggers you so much lol
32.
by maximedupre 1y ago
Yeah it could definitely be improved. For one I could use a better model than gpt-4.1, perhaps a reasoning model, but I didn't want to pay more for a free tool.
33.
by maximedupre 1y ago
ChatGPT pretty solid ngl haha
34.
by maximedupre 1y ago
Yes it assumes exactly that aha. It's also a free tool, so there is a tradeoff between which model I chose and much context I gave it :)
35.
by maximedupre 1y ago
How is it broken? You submitted two websites that don't exist anymore and you expect good results? It is LLM based so it responded with the information it had lol
36.
by maximedupre 1y ago
lmao
37.
by maximedupre 1y ago
It visits the website that you pass it, analyze the business, and then tries to find competitors. Sometimes it misses the mark though! Can you tell me which URL you used, I'd love to debug it to make it better :)
38.
Show HN: Competitor Finder – Paste your domain, get your top competitors(champsignal.com)
5 points by maximedupre 1y ago|16 comments
39.
Show HN: I built a app to monitor your competitors and get only high-signal info(champsignal.com)
1 points by maximedupre 1y ago|0 comments
40.
Show HN: Clip → Script(zeroface.ai)
1 points by maximedupre 1y ago|0 comments
41.
Show HN: I made a SaaS to generate viral scripts for faceless Shorts(zeroface.ai)
1 points by maximedupre 1y ago|0 comments
42.
Show HN: Ghibli faceless videos (and other niches)(zeroface.ai)
2 points by maximedupre 2y ago|1 comments
43.
Show HN: Free TikTok Voice Generator (Voice overs for short form content)(ttsvibes.com)
4 points by maximedupre 2y ago|0 comments
44.
by maximedupre 3y ago
Oh! That looks interesting :D I'll go through the article!
45.
by maximedupre 3y ago
Ty sir
46.
by maximedupre 3y ago
TY Nik :)
47.
by maximedupre 3y ago
Right! That makes sense. I guess you still need some technical knowledge though.
48.
by maximedupre 3y ago
Ty :) Yep, all your tests remain on your dashboard unless you delete them.
49.
by maximedupre 3y ago
I'd be curious to know what the other a/b testing tools are doing regarding this problem if anything.
50.
by maximedupre 3y ago
Hey Joseph, thanks for the comment. To be honest, this scenario never crossed my mind once xD. I have not implemented any solution to this problem... and as I'm writing this comment I'm not sure what the solution would look like.
51.
by maximedupre 3y ago
:D It's better if all your experiments are independent from each other however.
52.
Show HN: Splitzy – Dead simple and privacy-friendly website A/B testing(splitzy.io)
21 points by maximedupre 3y ago|18 comments
53.
by maximedupre 4y ago
Interesting project Falak! Keep it up
54.
by maximedupre 4y ago
This looks like a handy tool :D
55.
by maximedupre 4y ago
Thank you so much dude :D Onwards and upwards!
56.
by maximedupre 4y ago
Thanks Daniel, I guess xD That was quite cryptic, but I think I get it haha. Cheers!
57.
by maximedupre 4y ago
That's an interesting app :) What motivated you to build peerdiem?
58.
by maximedupre 4y ago
You understand Birdy's mechanism correctly :D > How do you make sure the same user who was on A variant yesterday doesn't get the B variant today? You can't. You make the point that the stats generated by Birdy are not 100
59.
by maximedupre 4y ago
I suppose that the issue is that a single user could see both versions, which could skew the data. I'm not sure I totally understand why, because if a user follows you after seeing the other profile version, it might be because he pref
60.
by maximedupre 4y ago
Thanks for the comment :D > There is no “testing” going on here, unfortunately. You’re just taking turns placing users in the treatment or control group, arbitrarily, depending on when they visit the page. How can you measure any treatme

More ›