Founder & CEO of Mostly Serious. Helping companies plan and build for the future of how work gets done.
Springfield, Missouri
Joined April 2007
- Tweets10.9K
- Following814
- Followers3.3K
- Likes40.2K
Jarad Johnson retweeted
Across our projects @GoogleDeepMind, we are witnessing a genuine phase transition in the pace of scientific discovery. You will see a glimpse of this in our releases in the next few weeks.
I find this approach more taxing than simply managing dedicated threads. The forced context juggling with one agent quickly becomes frustrating.
And the layers feel like token burning theater.
This could change as agents become better at knowing what requires human interaction, what doesn’t, and how to surface those in efficient ways.
I've changed how I work over the last few weeks.
I now have one lead agent for each area of focus, running all the time, and a chief of staff agent that oversees them. Each lead manages its own sub-agents that write code, test it, chase reviews, and coordinate with team members.
Everything I need to decide or review comes into one prioritized queue, a simple local web app. Each product review includes screenshots and a video demo of the feature. I clear the queue in batches, mostly by voice and keyboard shortcuts. The leads also bring back ideas I hadn't considered.
Working from one queue, instead of switching between dozens of chats, has cut a lot of mental overhead. The system keeps learning my preferences and distilling my judgment into a "Brian brain," so its recommendations and first drafts get better before they reach me. The goal is fewer corrections per change.
Every change still goes through the same human review and security checks. It is eliminating tasks, not human judgment.
What an exciting time to push the frontier of productivity, especially as we build the economy for agents (agentic finance). We should all keep pushing in this direction.
Jarad Johnson retweeted
working at a closed weight lab and having it on the resume will be the cultural equivalent of this moment in time of history.
with the amount of capital raised they will do evil. to achieve roi.
facebook was cool at first but do you remember the cold winter where people didn’t want to be associated that they worked their during the elections?
not being involved with producing closed weight tokens is a good moral stance that would like others to see.
look past the money. look at where we are going.
Jarad Johnson retweeted
Introducing Anatomy
Describe a machine in one line. Claude brings it to life.
- Interactive isometric wireframes you can rotate, inspect and explore
- WebGL shaders for fire, water, plasma and more
- Real 3D motion, with individual parts that rotate and move
Install it as an agent skill. Fully open source.
Had the opportunity to attend @every's Thesis conference yesterday. It was phenomenally well done.
They did so many things well.
- Short Sessions. They kept things moving so no one topic went on too long. Speakers had to distill their messages. It made every moment more valuable.
- Intentional Connection. From the energetic lobby to the thoughtful lunch seating, every aspect of the conference felt designed to bring people together and allow for serendipitous connection. I met so many interesting and smart people solving important problems.
- Frontier Problems. Too many tech conferences feature people who aren’t actually working at the frontier. Every speaker seemed at least a year ahead of the rest of the world, often much further.
- Practical & Applied. Another thing most conferences get wrong. Thesis had some theory but it was far more practical. What is a software factory? How are companies actually adopting AI and where are most failing? So much of the content was as reaffirming as it was insightful, and I found that incredibly valuable.
- Little Pleasures. There was always a water, (really good) coffee, LaCroix, or snack nearby. The food throughout the day was delicious, and the staff were kind and helpful.
- Carefully Crafted: The design and atmosphere supported all the other decisions, while living beautifully in the background.
Excellent work to their team. If they do it again next year and you can attend, I recommend it.
Jarad Johnson retweeted
Subagents are the biggest cause of AI failures
In most harnesses:
1. Main agent writes a summary of what it knows
2. Subagent starts from that summary only, and makes assumptions
3. Subagent works and sends back a new summary of what it did
Same flaw as compaction: all summaries are lossy. This is how your assistant makes emails you never wanted, or purchases what you told it just to get the price of
So I tried something new: checkpoint agents
The initial agent has complete context of what needs to be done. And that input is already cached
Fork the session where it is. Send the fork ahead to do the task and it will do exactly what the main agent would have done; their context is identical
What is even better is when the fork finishes, the cache of the main agent is kept warm by the subagent (they share the same prefix)
I measured it on a 137k-token session with Opus 5.5:
→ Main agent does it on its own: 67 s, $0.38, context +8-17k
→ Checkpoint: 117 s, $0.68, context +2k
→ Classic subagent: 123 s, $0.57, context +4-5k
But here is what I found: avoiding large input is even more important than keeping cache. And most subagent tasks were done better either in series with retrieval, or split up to be parallel over fresh sessions from the start
I do not think there is a place for subagent delegation at all. There should either be a super-agent that orchestrates and does no work, or a long running agent that can retrieve from previous sessions, not a worker that completes tasks partially and delegates the rest to other workers
Don't design harnesses based on our assumptions of human behavior
Second this. Best event I’ve been to this year. Phenomenally well done.
if there was an “ai conference” that you should have gone to this year, it isn’t aiengineer, it’s @every. these folks ooze class, style and with a intimate moments for a small curated attendee list who get together to ponder all things about how AI is changing business. do not miss their next event…. if you can get in…
Jarad Johnson retweeted
Bad policy, bad character make bad politics: erratic tariffs, impulsive Iran war, blanket pardons, apparent corruption, weaponized justice, ICE brutality, mountainous deficits, hiding Epstein, abandoning allies and coddling enemies. Hoping for a course correction, for the country and for my party.
Jarad Johnson retweeted
BREAKING: For the first time in more than 80 years, a Democrat will represent Springfield in the Missouri Senate. Betsy Fogle has defeated Republican Melanie Stinnett and Libertarian Cecil Ince in SD 30.
Read it: sgfcitizen.org/government/el…
Jarad Johnson retweeted
i've reproduced the results from the MIT blog post today on weight-space programming. but one additional result i have to add: this is a new, step-change emerging capability that's appeared in just the latest set of models.
Jarad Johnson retweeted
Introducing Honest Thumbnails
A Chrome Extension to unslop YouTube thumbnails with stills from the video. Brings back the early 2010s vibe of Youtube where not everything was about clickbait.
While looking into how @karpathy's HackerTube works, I learnt that YouTube has a CDN that provides still frames from the following chunks ¼, ½, ¾ of each video.
So as you can imagine, it was only a matter of time until Claude turned this into a Chrome Extension. Honestly, it took me longer to make the icon than it took Claude to write the code.
Open Source and MIT. Fork it, steal it, do whatever your heart wants with it.
github.com/851-labs/honest-t…
Weekend hack: HackerTube. I was annoyed with YouTube's recommendations so I had Claude filter down 43M comments over 20 years of HN to 6,454 videos that keep being recommended. *Significantly* better viewing experience, ~2 hours with Opus 5.5:
hackertube.org/
- Excellent recs without any logins (state is kept in local browser session)
- Thumbnail is a representative frame from middle of the video instead of the super duper A/B test hyper-optimized user uploads
- Underlying data from the comment analysis as one zip file, point your Claude at it and ask for your own version
- HackerTube Longs ✊
A bit more on the process:
I use Claude Code (desktop app for Mac), Opus 5.5 on Medium effort as my default. I only talk to Claude and have it do everything, the only "manual work" I helped with in tandem was in the browser (reminiscent of MenuGen): 1) I had to create some kind of API key from Google Cloud UI to query YouTube for video metadata, 2) I booted up a VPS box for serving and 3) I bought the domain name. Once the box was up I gave Claude the IP and told it update, secure, harden, deploy and optimize the code, and it did the whole thing.
Many awesome projects that would have normally taken weeks/months even can now be few-shotted for breakfast. What else are we not thinking of.
Jarad Johnson retweeted
Anthropic says in its amended S-1 that its internal estimates and expectations about FUTURE models indicate that it may "imminently" meet the definition if "dramatically accelerated progress attributable to AI".
Under the RSP, this means that Anthropic observes or expects 2x the rate of progress in AI capabilities vs. the rate Anthropic would expect and the fastest rate of extended progress observed over at least 3 model generations without significant AI contributions to AI R&D, plus it's plausible that this doubling is due to automation of research and/or engineering such that it's likely that there will be even greater acceleration.
Note that "2x the rate of progress" means that 2 years of progress would be compressed into 1 year. So, if Anthropic would have improved its models 9x in one year without AI-enabled acceleration, this threshold would be reached when Anthropic expects to be able to improve its models by 81x in one year (9^2).
I now have a custom YouTube site for a specific aspect of my work and another for parenting advice based on my kid’s ages and interests. Lovely idea. nitter.gq/karpathy/status/211698…
Also true for their Granola killer, Flow killer, Wordpress killer, n8n killer, etc., etc.
Some of it is that they take a lot of shots and keep what works. Some of it is a lack of focus after launch.
In businesses this can be super painful, like the current abrupt switch away from CustomGPTs, a concept that easily could have molded into Dots had they iterated instead of ignored it.
Max 5x → Max
Max 20x* → Ultra
Actually 10x.
Not a great look that they changed the name but still chose to be dishonest.
We’re renaming our Max plans:
Max 5x → Max
Max 20x → Ultra
Max is still $100/month, and Ultra is still $200. If you’re a subscriber, there’s nothing you need to do.
We’re also removing the 5-hour limit for verified Ultra accounts, starting today with most existing subscribers. Newer accounts keep a temporary limit until they’re verified, which helps us prevent fraud and abuse. You'll get an email when it's removed.
Weekly limits remain unchanged. View our plans here: claude.com/pricing
Jarad Johnson retweeted
ChatGPT web app now has a new hidden shared usage limit checkout disclaimer - "Starting in December, plan usage will have a shared limit across all of ChatGPT and Codex"
Made the mistake of upgrading to the $500/mo ChatGPT plan before my $200/mo 20x usage expired.
The $500/mo plan is draining so fast. OpenAI is in huge trouble in a few days.
They either have an answer to Opus 5.5 coming tomorrow or they'll have to revert back on the usage cuts.
They did the thing. nitter.gq/thsottiaux/status/2115…
Day 25/
We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today.
We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.
And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.