Dave Swift
946 posts

Dave Swift
@daveswift
Software Enthusiast: https://t.co/KdjRiXDICq | Minnesota Twins fan | Husband & father of 4
Eden Prairie, MN Katılım Aralık 2007
307 Takip Edilen2K Takipçiler


@JohnONolan @Ghost I built most of daveswift.com by talking to Claude code to customize a theme and edit things via the API. The CLI will be a huge upgrade. Can’t wait!
English

Recently, I've found my preferred way to interact with lots of products is getting Claude to use a CLI tool, and then just talking to it out loud about what I want to happen.
So I made a CLI tool for @Ghost, to see what it would be like in our own product - it's called `ghst` and it essentially makes everything in the UI available by CLI (and MCP)
Mostly this was an exercise in doing it just to see if I could - but it turned out to be a lot more interesting than I expected
We've spent 10+ years focusing on having a clean, well designed interface for Ghost. It's something we care a lot about, and spend a lot of time on.
But within about ~1hr of using Ghost via Claude/CLI, it was hard to imagine going back to caveman-clicking around a browser to get something done. Particularly for complex or compound tasks that might require visiting several different areas of the app.
Which is both obvious, and at the same time kind of jarring.
I know Ghost's UI extremely well, and know exactly where to go and what to click to do the thing I want -- and even for me, using Claude is significantly faster/easier than clicking myself.
So how big would the delta be for a regular user who *doesn't* already know the UI inside out?
My initial thought was "huh, I wonder if UI even matters anymore?" - maybe everything just becomes a CLI / voice interface for a database, as various people have been suggesting about CRM tools.
But I don't think that's quite right. I notice when I interact with the product via Claude - I usually still keep the UI open, but my relationship to it is different. I use UI to see what happened, verify things look "right", and get an overview of what's going on.
Which is kind of familiar, because "agent does the actions for me / I review the results" has obvious parallels to AI coding workflows.
Before I'd be in VS Code all day doing the thing myself, but now I use Codex Desktop which is a *UI* designed entirely around optimising for: agent does the actions for me... I review the results.
Anyway, I don't know what my conclusion is here other than to say that AI+CLI is a really cool pattern, and I think it's likely to meaningfully influence what "UI" means over the next few years.
There are still tons of rough edges and reasons for why this is not yet a fully-formed paradigm (regular humans do not, and should not, ever need to know what "CLI" or "MCP" even means), but I like where it's going.
If you want to try ghst CLI - it's linked in first reply below
👇

English

@mcuban @adcock_brett @tbpn Rebuilding or building new human spaces that fit robots - that transition will take at least 50+ years.
English

.@mcuban says humanoid robots won't last more than 5-10 years.
Instead, we'll "design the house to fit the robot, and design the robot to fit the house."
"You could create a house where the pantry, the refrigerator, and the washing machines were hidden behind the garage, if you even have a garage. That way you could redesign the house so that all the living space was for people."
English

@ogandreakiro @nottecore It was ‘look up the stock price of APPL on Yahoo Finance’
English

@daveswift @nottecore @daveswift Hey! Mind sharing which prompt you've tested? I'll look into it and check what went wrong. Thanks for testing + reporting!!
English

Anything API is live on Product Hunt! 🔥🚀
Most websites don't have public APIs. Anything API fills that gap.
Describe the browser work you need. Our agent builds it, deploys it, and hands you a callable endpoint that you or Claude can invoke from everywhere.
Any website. We deliver the API.
producthunt.com/products/notte…

English

@SawyerMerritt Guess people only thought they were ugly when they couldn't afford them. 🤷♂️
English

@nevmed The anker cube is great too. 3 in 1 also, looks great on a nightstand and super portable for travel. a.co/d/06EXh7Mz
English

This is kind of a random product, but I've re-ordered this like 4 times (one for all the rooms) and still need more!! I also gave them away as White Elephant gifts during Christmas and people loved/used them.
It's a simple magnetic phone charge that can charge an iPhone, iWatch, and AirPods at the same time.
I also have to mention I LOVE how they sold me the product through the images alone! I didn't even really read the copy since the images alone explained it so well:
The feature I love the most is being able to just magnetically plop my phone on it and use it as a little baby monitor or for my workout app...and it's also charging and able to swivel:




English

@noahkagan Confession... I think about killing my agent all the time. Someday I'm going to do it.
English

Open claw is still overrated.
Here's my hot take:
Everyone asks around how do you use it - why? Cause no one has a use case they find invaluable besides making dashboards or trying to arbitrage Polymarket.
Maintenance - I spend 80% of my time just keeping it online, remembering or fixing things. It forgets time and time again. Also sucks up all computer resources regularly.
All the posts about SEO optimization, how they have 15 AI employees, etc. Are from people not making money.
Token costs > Executive Assistant cost ($50 / hour). Turns out using better models and running tasks around the clock costs money. And trying to debug or explain things takes way longer than sending to my assistant (for now).
It hurt my X account. When I had it run my X and then check my X for stats - I got throttled since it looks like a bot (cause it IS a bot!) - didnt realize it for a week.
Every-single-person who's bragging about OpenClaw is mostly lying. When you ask them how it's really going it isn't as they seem.
Is it awesome and hugely potential, yes. Still someways to go...
English

@Jason Start self hosting this. Why would you want to pay Slack even more just to access your own data via an API Check @Mattermost.
English

Introducing Rork Companion Mac app
Until now, if you wanted to build apps with Swift, you had to install 30GB of Xcode, struggle with certificates, TestFlight, and pay $120 to test on your iPhone
Rork Companion solves all of it with a Mac app that can install up to 3 iOS apps on your phone, no $99 Apple Developer account needed
P.S. You still need an Apple Developer account to publish to the App Store
Rork@rork
Introducing Rork Max AI that one-shots almost any app for iPhone, Watch, iPad, TV & Vision Pro. Even Pokémon Go with AR & 3D. Max is a website that replaces Xcode. Install on device in 1 click. Publish to App Store in 2 clicks. Powered by Swift, Claude Code & Opus 4.6.
English

Out of the box, AI Tools are dull axes, more likely to cut your foot off than make a perfectly stacked cord of firewood. They have to be sharpened, tuned, and maintained to increase productivity. That takes labor. Why don't we have self-driving cars? Error rate is too high, can't be trusted. It can't even be explained, because models are ultimately opaque.
English

🦔 Jason Calacanis says his company hit $300/day per agent using Claude's API at only 10-20% capacity, which scales to around $100,000/year per agent. Chamath Palihapitiya added that he's now asking "what's the token budget for our best devs?" and said AI-assisted developers need to be at least 2x as productive just to justify the cost. He said this is actively happening inside his company or he'll run out of money.
My Take
This was always the obvious trajectory. AI providers subsidized usage to drive adoption, and now the subsidies are ending. The consumer plans are likely loss leaders subsidized by VC money, and the gap between what individuals pay and what it actually costs to run these models is closing fast.
I'm struck by the surprise from people who should know better. These are sophisticated tech investors just now realizing that running agents 24/7 burns through tokens at rates that dwarf human salaries. A human engineer runs on coffee, remembers context from years ago, builds institutional knowledge, and doesn't rack up exponential costs the longer they think about a problem.
Agents waste tokens constantly, researching and validating things that don't need validation, spinning up subagents for simple tasks when a straightforward approach would work fine. The companies that fired engineers to replace them with AI agents are learning that you can't negotiate with an API bill the way you can renegotiate a salary. And unlike employees who might stick around during a rough patch, the meter just keeps running.
Hedgie🤗
English











