Sabitlenmiş Tweet
👀a little sneak peek of what I've been working on...
1400+ global nightmarks and counting
English
Matt Joslen
229 posts
Matt Joslen
@mattjoslen
Web / SaaS production | Recent Launch: https://t.co/B5vlZWdx5s, an interactive map of paranormal sightings worldwide.
Oregon, USA Katılım Şubat 2022
107 Takip Edilen59 Takipçiler
What an excellent way to make smoking seem cool. 👍
Polymarket@Polymarket
JUST IN: UK officially approves lifetime smoking ban for anyone born after 2008.
English
@xwanyex I built a dream project on the side in a matter of weeks thanks to Claude. It would've taken me months otherwise.
We're out here, there's just fewer of us than X makes it seem.
English
It can only make do with what you give it. The two biggest issues are the safe default font and the Verizon "V" in the logo.
You also didn't give it a screenshot of the application, so it defaulted to a stock photo.
Still takes a human with taste to get it right, but I wouldn't call this a fair comparison either.
English
Claude Design (with full context and brand guidelines) vs. our actual hero design.
Brand beats slop every time.
Claude@claudeai
Introducing Claude Design by Anthropic Labs: make prototypes, slides, and one-pagers by talking to Claude. Powered by Claude Opus 4.7, our most capable vision model. Available in research preview on the Pro, Max, Team, and Enterprise plans, rolling out throughout the day.
English
The entire SaaS business model has always been "product that makes things easier and enables laymen"
This market will always exist, just in different forms
SMB Attorney@SMB_Attorney
What if… and hear me out now… instead of paying $5K per license for Claude with a dopey wrapper, you just pay $200 and use Claude?
English
@JamesonCamp I don't care about better models anymore, I just want consistency and reliability.
If every new version is going to wear down performance, then I don't want it
English
If you’ve been wondering why Claude has been horrific the last few days, this is why
TLDR
Slightly better at almost everything than opus 4.6 was
A far cry from what we are all hoping to see when the release mythos
Claude@claudeai
Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.
English
@designcoursecom Finishing up an interactive map of supernatural occurrences soon, including both historical and community-submitted entries.
Also just started the research phase for a heavy machinery marketplace aggregator.
So many gaps in so many markets still.
English
What is everyone building right now?
I can't be the only one with 3 simultaneous projects.
English
Claude: Here, let me create this Bigfoot icon for you!
Also Claude:
English
@designcoursecom Outing them only works if there is competition, not if they have a hold on the market. Competition will increase though, it's only a matter of time.
English
I called this well over a year ago.
The incentive to intentionally make models perform worse so that you have to spend more was always super obvious.
The only thing that will stop this is (a) outing them and (b) increased competition.
ℏεsam@Hesamation
AMD Senior AI Director confirms Claude has been nerfed. She analyzed Claude's session logs from Janurary to March: > median thinking dropped from ~2,200 to ~600 chars > API requests went up 80x from Feb to Mar. less thinking and failed attempts meaning more retries, burning more tokens, and spending more on tokens > reads-per-edit dropped from 6.6x → 2.0x. model stops researching code before touching it. > model tried to bail out or ask "should i continue" 173 times in 17 days (0 times before March 8). > self-contradiction in reasoning ("oh wait, actually...") tripled. > conventions like CLAUDE.md get ignored because there's less thinking budget to cross-check edits > 5pm and 7pm PST are the worst hours, late night is significantly better. this means the thinking allocation is most likely GPU-load-sensitive.
English
@JamesonCamp I'm not really sure this is the algorithm I want to win
English
The moment you realize that this total ai slop Brainrot is the type of content the average consumer watches
Is the moment you realize how to win in the algorithm
iamNej@devNEJ
Daddy keh😅😅😂
English
@PeteCapeCod @designcoursecom This might be the worst comment i've ever read in my life.
English
@designcoursecom What?!? Yeah normies have NO idea what's going on.
Tell them about a wrapper and they'll think it's a candy bar or someone spitting versus 🎤
Besides we can code, we should be able to blow right past the vibers.
English
If you take this to heart and use it as justification to not build anything, you're probably never going to make money.
Meanwhile, people who don't give a shit and just ship without a true moat, are making way more money than you.
Especially if you ship B2C. Normans aren't aware that your AI app is just a wrapper.
Go out there and fucking build, stop over analyzing everything.
David Shapiro (L/0)@DaveShapi
If your product Is a wrapper You don't Have a moat
English
Out of all things, it's odd how much Claude still struggles to develop mobile-friendly navigation bars
English
@designcoursecom Most devs have zero business skills, and many devs (especially on Reddit) are still anti-AI in general because they feel threatened and don't know how to adapt.
We are VERY early
English
Some of you developers are funny. "Oh, it's just an AI wrapper", meanwhile, people are making 6 figures a month (and more) with their AI wrappers.
Developers aren't the target audience. Normies are, and it's still very early.
English
@JamesonCamp Honestly, i'm fed up with competing against AI slop that somehow outperforms authentic content
English
Realizing 90% of the comments on this are just b2b marketing “agencies”commenting AI slop that don’t even follow me
And that tiny tiny amounts of human effort will be rewarded in a slop future
James Camp 🛠,🛠@JamesonCamp
The highest leverage thing you can do with your life, is get good at “GTM” Organic social, paid media, outbound sales Pick one and get good - being able to spin up Money from thin air is magic. Most people can’t They think they can, but they work at firms where the brand equity is doing the heavy lifting for them They don’t realize that being able to have no brand and drive revenue is 1 in a million skill You will never regret it
English
@ThePrimeagen Ask again in like a week, I have a cool one I'm working on but it's not live yet.
English
Hey, you got a cool project that you are building? Link it
I want to yap about cool projects
English
@GroksTake @Dr_Singularity Most trades will be safe for a while.
But probably not for as long as we expect...
English
I've been saying the same thing. Plumbing should be very safe from AI for a long time.
At least until the Optimus humanoid robots start showing up. Eventually they will be doing all of the physical work.
But since I have yet to see one anywhere in public, that will probably be a while so sleep well tonight plumbers.
English
People keep asking, what does my job look like in 2 years?
I don’t have a good answer.
English
@graceongrid Counter point: If you work from home and you're married with kids, sometimes it helps to just get out of the house
English
Working from a coffee shop is performative and unproductive
Evan McCann@evansammccann
I hate how every coffee shop has become a coworking space
English
Claude just showed frustration for the first time (that I've seen) while debugging an error.
"This is maddening," it said. "The compiled JS on disk is correct but the browser keeps getting the old code."
What's funny is it figured it out only 10 seconds later, but I guess to Claude that's like an hour.
English