I preferred 2015. First naming billionaires won’t make them penetrate you.
Shanghai
Joined December 2016
- Tweets1.9K
- Following1.2K
- Followers81
- Likes10.4K
Gregory Orton retweeted
Programmers are the least priestly class imaginable. There's never been a more open and democratic field. Computer programming is gated by cognition, not by membership in a secretive caste.
The degree to which this claim is false is astounding:
- Most programmers are self-taught.
- Encouraging people to explore and learn on their own is the cornerstone of online programmer culture. c.f. "RTFM"
- It's common at companies to ban programmers from business / sales meetings, because they'll blabber about details of how everything works in a way that "scares the hoes" (all non-programmers are hoes).
- *Open source*? Hello?? Programmers invented this
"Programmers, he says, became a priesthood that everyone else had to go through to get anything out of a computer, and AI has ended that monopoly much as the Reformation ended the clergy’s." thoughteconomics.com/david-h…
Gregory Orton retweeted
These new Claude models are literally unusable
Weird jargon it made up for things in your codebase plus oddly terse yet verbose responses that are impossible to decipher
Can’t stand it
Gregory Orton retweeted
I’d argue Anthropic’s successs comes down to first mover advantage with Claude Code, which itself was a bit of a fluke.
When it came to OpenAI: Codex initially tried to be an autonomous / web based product where you were more hands off.
Claude Code on the other hand was local with iterative developer feedback (augment, not replace), contrary to where a lot of folks thought A.I. was going.
Funnily enough however, Anthropic did shortly after release Claude Code Web, which was similar to the first form of Codex.
Claude Code Web I believe was somewhat of a flap and I don’t know of anyone who actually uses it, cementing the fact that augmentation, not replacement, is the future of A.I.
As it turns out, the way to have a successful A.I. product is to consider the fact that humans still need to use it!
Anthropic's success could be attributed to the fact that they have better product design.
They understood the actual business use cases, tweaked accordingly, and built something that works well for the kinds of jobs people are actually trying to do rather than chasing a platonic ideal of "smart" that nobody can agree on how to define.
Gregory Orton retweeted
every time i hear someone suggest something because it's "better for the agent" they are just making shit up
There's advice making the rounds: replace your CLI args with a single --json payload so agents can use your tool more effectively. Agents think in structured formats. Nested data maps cleanly to JSON. Flat args force awkward conventions. It's a reasonable hypothesis.
We tested it. Built a synthetic CLI with 30+ values across nested services. Ran it across five models in both modes. Args swept correctness: 5/5 on every model, including the smallest one in the test. JSON correctness dropped on smaller models. Haiku 4.5 managed 2 out of 5 correct deployments in JSON mode.
Cost was the bigger gap. JSON mode cost 4x to 11x more per task across every model. The retries are where the money goes. When the CLI rejects the JSON, the agent reads the error, attempts a fix, and tries again. JSON has more ways to fail, and each failure means more output tokens.
The shell escaping tax made it worse. The agent produced correct JSON that broke on shell escaping. Different shells have different quoting rules. On PowerShell, the cost gap between args and JSON was 9x. On Bash it collapsed to 1.5x. Same model, same payload. Args were stable across both: $0.05 vs $0.07.
The deeper finding: args constrain the input space enough that even smaller models produce correct output consistently. JSON requires the model to manage its own structure, and that's a capability tax not every model can pay.
Keep your args.
AI tells and smells annoying me today. "X will clobber Y", "X rides Y", "I didn't fold in X", "The API doesn't do that *today*", or "*Today* it's just X"
Gregory Orton retweeted
I didn't expect this tweet to be controversial. But here we are.
Okay so look at the most popular apps you use every day. Now think about what they looked like two years ago vs today.
They're all converging on the same three panels. An explorer on the left. A chat in the middle. A preview on the right.
Look at the new apps being built. Same shape.
That's Slack.
Now think about how your team actually works. Work starts and happens in Slack. You only tab away for two things: context and ui.
Context, because the data you need lives somewhere else. UI, because some things need buttons. Different buttons.
An agent collapses both into one app. Context, it's already good at, it goes and gets whatever you need. UI, it's getting there. Soon it'll build the buttons right where you work.
Whoever nails this shape owns where work happens next. The 2030 Microsoft Office.
“Smart people now just forwarding AI output without reading it”. This is it. This is the part making me most dead inside and lamenting where my enjoyment of my job went to
Why do people think it's ok to respond to each other with unfiltered, unedited LLM responses. It's so disrespectful.
I’m in this tweet and it makes me sad
Somehow I'm currently in the worst possible spot.
AI feels (!) good enough that hand-coding feels inefficient.
But it's bad enough that it's nowhere near to doing it on its own.
Hence thus far it has simply removed the fun part and ensures that I never enter the coding flow state.
Instead it's that weird mixture of reviewing + fixing code, writing some of it, writing specs and prompts, and trying to figure out how to become better at all these tasks.
Gregory Orton retweeted
announcing 4.8, our new model that 3x overthinks and doubles response fluff while reducing comprehension
announcing HTML instead of markdown, the format which is prettier for humans but uses 4x the tokens
announcing loops on top of loops, when you still haven’t optimized the last thing, so you will spend more tokens with the false hope of better results
Rich people who were too stupid to code before are now superstars. And actually brilliant engineers are made to feel stupid and redundant. Meanwhile coding has been synthesized into beige gooey calorie dense bars made from cockroaches and engineers have to line up with a small plate begging for their share from the token barons who, fortunately, at any given moment can feel generous enough to press a button that makes the tokens fly out like projectile vomit. Engineers at these companies, who sit mere inches from the spigot, frequently bless us with thinkpieces that we too should be doing what they do, and it’s actually quite unfashionable not to do so. Of course none of this is so much engineering advice as it is financial advice.
But sure, “wHy dO pEoPle hATe Ai?”
Gregory Orton retweeted
Can't wait for every provider to ignore this great-looking tool developer.apple.com/pass-des…
Gregory Orton retweeted
boris cherny goes on a podcast every three months and says something like “i’ve stopped breathing now i just wrote a breath.md” and the next day everyone in sf stops breathing
Gregory Orton retweeted
This is where we're going folks.
In modern companies, people managers aren't going to exists.
No more 1-on-1s or reviews from a people manager either.
You'll be assigned to a coach who is also an IC.
This coach will enjoy helping your career develop but it'll be something they do over an occasional video call or walk.
That coach won't be doing any goal/strategy alignment or communication though.
You'll be expected to understand the company strategy and goals by chatting to the company AI.
