prompt engineer @browsercompany | I make apps. My code runs on 100mil+ phones | former iOSdev @Twitter | indie dev @mindGoblinStdio
Seattle, WA
Joined October 2012
- Tweets47.2K
- Following9.4K
- Followers55.4K
- Likes446K
New Jev like model from Cloudflare!
Decision models are out already lol.
We got system 1 models now
Cool thing about this, unlike Jev which can only return decision probabilities, this can also extract text, computed and estimated values. Because it's not generating new text tokens, it can't hallucinate, but it still lets you pull exact words and figures out into structured json.
Super handy for asking questions like how many, or how often. A new type of smart if statement similar to Jev but for different questions.
a brand new model paradigm! introducing 𝚌𝚕𝚊𝚒𝚛𝚒𝚌: a system one model for extracting, computing, estimating, and predicting values from unstructured inputs.
blog.cloudflare.com/clairic-…
People wanna hate on this but I do think 2-3 days, staying up late, of extended focused pour your soul and mind into the computer and do almost nothing else is the ideal coding flow now.
Unfortunately most jobs (and most people) are not setup to handle this or practiced with this intense on/off cycle, so it’s sounds and looks crazy. Not for everyone.
But like the standard tech co office hours of 9-5 or more, 5 days a week with crazy amounts of context switching and meetings is an insane working style too and produces many undesirable psychological effects. Neither of these approaches are without health concerns nor downsides.
Worst part of my job by far is the way hours are imposed as a regular low grade chronic stress grind and not these intense sessions.
I would prefer a more manic and obsessive work style personally.
It genuinely takes hours to get into a full proper vibecoding flow state. Similar to how meditating 20 minutes a day doesn’t really kick in until day 10 or so. Takes time for neurons to change and your body to get there. By hour 12 of vibecoding you reach areas of thought and creativity you can’t reach otherwise.
Alas the plight of the creative and the obsessed. People see out of distribution individuals and yet don’t seem to wonder if their quirks and flaws are what lead to their successes in other areas in the first place.
If proactive agents become the standard, then you need start creating an info diet for them.
I be sending my agent interesting tweets and papers with no instructions or follow ups, just check this out and summarize, so the agent has good seeds and knows the interesting stuff and cutting edge trends so that later on when we brainstorm it can tap the neat stuff in its memory from 4 weeks ago
Nick Dobos retweeted
You need to prompt with love in order to guide large agent swarms
I’m not kidding. 100% serious & literal.
Values, culture, feeling, and story based prompting is the only way to steer 10000000000 decisions a second. Same system CEOs use scale decisions.
Statusmaxxer vs actual progresschad
Ok that phrase is now seared into my brain. Once you see the larpers it becomes so apparent who’s got the sauce and who doesn’t.
True, AI = better games, BUT I have no idea people are gonna monetize this.
Business model of gaming is broken not the technology.
I’m trying to be optimistic cuz I do enjoy games, but when the gap between creator and player vanishes, when fan fiction modding with agents becomes the primary way you play, players will create more than the authors.
When that happens modern gaming business is going to evaporate.
The golden age of journalism is social media, where everyone can report. Notice how well those businesses are doing? The papers and publishers are dead. We literally call them legacy news media lol. But more people than ever now make a living from reporting, writing, and entertaining on writing platforms and social media, and these new platforms are the most valuable in the world.
But “gaming” the way we think of it from the last 30 or so years is dead.
That doesn’t mean people won’t play games or make money, but the types of games and the business models that enable them both to work, and for people to spend time on them, is going to be so different.
Ngmi.
If you’re not speeding up and doing more becuase of AI, you won’t have a job soon.
Ask for more work?
“We’re gathering requirements”
You’re fucked bro. You are now the person who needs to be proactively problem solving and figuring out what to build. That’s the role.
There isn’t enough work?
…
Are you an idiot? This is impossible. There is always more to do. Your business has goals… Make your boss, your team and yourself more money. If you can’t figure out how to do that. Again, you won’t have a job soon.
Apple is banning feature switches.
Lmfao.
Every large app with a big dev team has hundreds if not thousands of these configuration switches to manage rollouts and testing.
Which probably should be reviewed. But previously Apple let people simply ship “bug fixes and improvements” and didn’t bother reviewing anything. Completely ignoring how modern companies would roll features out to 10% of users as tests before rolling out to everyone.
But now, oops that WIP feature you have turned off? Or only for a small user set for testing? Illegal. You are banned.
Apple is so so far behind current dev standards it’s wild.
Apple can now delete your whole developer account for a feature hidden in your binary. On Friday it became a separate rule, 2.5.19.
Apple already rejected one of my apps for this. I'd left three testing switches in the release build: screenshots-mode, slow-entitlement, UnlockedMockPurchases. Nobody could reach them from the App Store, but Apple wrote that they "appear to have been intentionally hidden".
Now I run strings on every archive before I upload.
ALT App Store Connect message from Apple on 2026-09-08 about Snowball Debt Payoff Planner: Guideline 5.6, Developer Code of Conduct. Issue description: We've identified a pattern of unusual behavior with the app that is commonly associated with fraudulent activity. Specifically, the app contains features that appear to have been intentionally hidden during the review process. Manipulative and misleading behavior is not consistent with the guideline 5.6.
Explained
1. Biggest problem with op’s charts is the assumption that there is no negative effect. Teachers always seem to forget it’s possible to teach people the wrong thing and have them end up traumatized with infohazards and maladaptive behaviors and thoughts. There’s not 0 learning there is backwards learning.
2. The original chart shows smart people being pulled back. That’s only true if AI is dumber or less creative or less able to execute “taste”. Which breaks if AI advances in capabilities, when AI has already proven to be superhuman in math, coding, and other disciplines, at minimum it’s able to do some things better, and likely more. So the spike is in the wrong spot. Few if any people are or will be pulled back in the long run.
3. The original charts also assumes AI users have 0 input or control over the final quality of the AI output. Ridiculous assumption. You are the one typing prompts and asking questions. You are the one picking which AIs to pay for, chat with, and what settings to use. You pick and edit the final work. You are the one aiming it! Your inputs matter!
Smart questions and more experience on how to use AI tools gets you better results!
This effect is recursive as you learn & accelerate with Ai, or downward spiral.
—
What’s actually going to happen:
Acceleration in every direction.
Smart people get smarter.
Dumb people get dumber.
Intelligence Gap widens
it doesn’t shrink
Every human emotion we experience will be dialed up to 11 and amplified with the power of the machine. The same way social media and internet communities often push and create extreme niche behavior.
Those who aim it properly will go far.
Those who don’t? Well. At least we have safety classifiers.
Standing out might actually be easier in an AI filled world, because people will be 10x more expressive due to increased leverage & ability to automate and do things.
For example, previously to run a burger restaurant or build an app, you needed lots of people.
Now with robots and ai coding tools one person can do that, which means we get 10x as many unique takes on it. So to avoid competition these businesses are literally forced to create variation. We get 10 different burger shops, 10 different apps, all doing their unique twist.
Abundance increases competition, diversity & quality.
A lot of people are stuck in this thinking that volume = factory = same. Cookie cutter everything. Housing blocks & misery.
But that’s no longer the case.
Abundance = We now can have 10 factories all making unique things.
