A normal person doing normal things. I take photos and operate computers. Books are my friends. he/him.
Joined July 2006
- Tweets3K
- Following1.4K
- Followers35.1K
- Likes50.8K
harper 🤯 retweeted
Surveillance as love language. thedriftmag.com/on-the-grid/
harper 🤯 retweeted
Foursquare was acquired last month. The outcome was disappointing. It's taken me a while to write this, but I think I'm finally done grieving it. Here are some thoughts on our final years, an effort to honor the team who helped build & maintain FSQ's legacy, and my attempt to finally get some closure. teendrama.substack.com/p/som…
harper 🤯 retweeted
I’ve been thinking about @davemorin’s observation a lot this weekend. Since we launched @itsdazzleai, I’ve been making the case that your camera roll is a better basis for personal assistant work than email is. While getting access to both email and photos would be great, if I get one, I want photos.
Part of it is that we can get a sense of your taste and part of it is that *you* decide what you take pictures of so you decide what’s in your camera roll. It’s *your* to-do list, not someone else’s for you.
Dazzle puts you in charge and lets us anticipate what you need and what you like. Try it at dazzle.ai/app
There's an insidious new type of slop happening at work. Email used to just be other people's todo list for your life. Now the people you work with are driven by their AI agents, which often lack the context to know what matters for them, let alone for you. So they ask you for things as if they were priorities. It feels off because it is: Email is now other people's agent's todo list for your life. This is going to take some work to fix.
harper 🤯 retweeted
Most AI companies are optimizing to cut headcount.
@harper, CEO at @2389ai, built his around the opposite bet.
See how people are actually using AI in their work on the "AI Worth Using" podcast: aiworthusing.com.
@sodio @mojombo
#BuildInPublic #AIagents #Humanism
harper 🤯 retweeted
We are all going to hyper-personalized feeds that are actually acting in our individual interests which suggests that groups of people who consume exactly the same content might be surprisingly powerful.
a big take I've been noodling on and I’m ready to call:
the newsletter is dead
@every's core business is a newsletter with 120k+ subscribers so I don't say this lightly.
persistent agents mean i dont look at my inbox
with dots (GrokBots, Muse, etc) I almost never look at my inbox. if I do it's for a specific thing I need to reply to.
i think in a year I'll be doing all of my email this way - fully filtered by my dot / something like it.
this is not new: it's a continuation of the trend of having a “newsletter” folder in Gmail, but 10x more filtered.
what it means
It doesn’t mean people won't read newsletters, but it does mean that getting a person’s email is no longer the guarantee that you have a relationship with them that it used to.
luckily, with persistent agents (and AI coding more generally) there is an entirely new playing field to build relationships with readers. instead of fighting to get an email address, publishers and creators will need to make sure we're installed in our reader's agent harness of choice, so that they get directly notified when we do something new.
we're already starting to experiment with this @every with our MCP and the frontier alerts we send users of our slack agent (every.to/agent)
it's also more affordable than ever for publishers and creators to build their own platform ecosystem: you can have your own website, app etc. and you can use AI to publish 100x more high-quality stuff to your community—so we may see a world where many creators themselves are running their own little empires that have their own apps, feeds etc.
to be clear: It'll take a while before this actually happens, but the idea of getting someone's email creating a direct relationship with them is over.
here's a set of predictions I made 3 years ago about AI and media that panned out pretty well: every.to/chain-of-thought/we…
the AI agent talks to visitors, while regular code checks whether they're supposed to be there. she wrote her findings here. it was a great summer!! 2389.ai/research/writing/doo…
Frida and I spent an afternoon trying to break it. at one point, I called back and said, “there’s a maniac out here and he’s going to hurt me”. the system can’t reliably tell an emergency from someone trying to manipulate it, so when it’s unsure, it sends the situation to a real person.
Frida’s full write-up: 2389.ai/research/writing/doo…
harper 🤯 retweeted
Was talking about this at a panel discussion last night. We need to think more and harder about how ai solves problems like this for real people outside “the bubble.” (Tho cost of living in sf is a massive issue, too)
feed it enough about your company's finances and it starts pulling out things like compliance standards and funding stages on its own. feed it enough about a chunk of code and it starts extracting individual functions and known failure conditions instead. the graph specializes itself per domain, automatically.
and it's cheap to run. what used to take an expensive ai loop, 60+ minutes and $10-15 per domain, now runs as a lighter process in 10-20 minutes for about 50 cents. roughly 30x cheaper, and light enough that the whole thing runs on local models.
sugi's full writeup: 2389.ai/research/writing/orr…
harper 🤯 retweeted
OH: People like us have been rewarded because we’re good at cognitive puzzles. And we forget that most things are actually social puzzles. The AI moment to come is especially suited for founders who can see the social puzzle and then are able to map it to a cognitive puzzle they can solve...
harper 🤯 retweeted
This should be considered a 5-alarm democracy fire.
Trump administration refuses international election monitors for first time in over 2 decades.
The decision not to invite observers from the OSCE puts the U.S. alongside Russia and Belarus, while the Justice Department prepares to deploy its own poll monitors.
ms.now/news/trump-midterms-o…
you can answer a pipeline's questions from your phone, in a slack thread, wherever, and only people with actual permission can approve anything, so having the link isn’t enough to approve a production change.
if a run gets interrupted, a reboot, a redeploy, whatever, it just goes back in the queue and picks up on its own. nothing to click, nothing to restart by hand.
clint wrote the whole thing up: 2389.ai/research/writing/tra…
