@fchollet My favorite analysis of this subject in terms of personal and political organizational dynamics is Robert Kegan’s _Immunity to Change_.
English
Ray Deck @ statechange.ai
6.3K posts
Ray Deck @ statechange.ai
@ray_deck
Founder @StateChangeAI. Youtube: https://t.co/6PBjwzkS4q Statechange: https://t.co/pxiIlMM5TD.
East Burke, VT Katılım Temmuz 2012
1.2K Takip Edilen890 Takipçiler
Most documented psychological biases are not irrational, they are highly optimized, energy-efficient shortcuts meant for a biological substrate operating under strict real-time physical constraints and a limited caloric budget
English
I don't know if i'm doing it right yet, but I find that the more I use AI, the more I'm pushed to _only_ do the hard work, making it a much bigger part of my day
Gergely Orosz@GergelyOrosz
I find myself doing a lot better work, being more satisfied, and also learn a lot more+faster when I do *the hard work* and don’t outsource it to AI. As in, I’ll use AI as a *tool* with substasks, additional research: but I don’t turn off my brain or kick back, assuming it can do the work for me. Every time I “hand over the” hard work part to AI and mentally turn off, I either regret it or find myself eventually needing to go back and spend more time on it. I also see slop work coming out from people who assume the AI does better work than they would.
English
Ray Deck @ statechange.ai retweetledi
Joined a new AI-native company this week and it’s kind of wild how different it feels already.
The laptop arrived, I logged in, and an agent basically took over from there. It set up my dev env, pulled repos, fixed dependency issues, got permissions approved, pointed me at the backlog, linked the architecture docs, and surfaced the Slack debates I actually needed to read before touching production.
When I needed context on something, I asked the agent and it found the exact thread from months ago explaining why a decision was made, who owned it, the related Linear issues, and the PRs connected to it.
I’ve only been here 3 days but it honestly feels like I’ve worked here for a year because the usual friction and scavenger hunt for context just isn’t there anymore.
We should probably stop calling this “onboarding” and rename it to “mounting” because this feels a lot more like mounting a distributed filesystem called “institutional memory” than slowly getting drip-fed context over 6 months.
English
Ray Deck @ statechange.ai retweetledi
I am noticing a really interesting trend:
my friends who are not technical are getting much further with open claw (etc), etc than my friends who are technical.
it seems that the ability for us to imagine the limits is artificially causing limits that are not there for folks who are not imagining limits.
have you seen the same thing?
English
Ray Deck @ statechange.ai retweetledi
The quantity of code that devs ship has roughly 10xed. But net developer productivity (value created by unit of time) is only up by a bit, if at all. Part of it is that the additional code is solving more incremental problems. A bigger part is that the new code is creating problems of its own.
English
@WindAddict I want to see windaddictMCP! John as a super high end tool call.
English
@ray_deck lol. GPs: for $20,000 I will actually read decks with my eyes and report back if your AI has been fooled. I have excellent reading comprehension skills.
English
This is happening a lot!
Karri Saarinen@karrisaarinen
What if we built better software, not just more of it?
English
Ray Deck @ statechange.ai retweetledi
the nuance to making this work is you need a dramatic efficiency gain in a narrow, repeatable area - say a 90% improvement for 5% of the operating surface, and then you expand from there
the failure pattern from the "tech enabled services" era was modest efficiency improvements across the entire organization, offset by the difficulty of a software founder running a traditional business
Deirdre Bosa@dee_bosa
"SaaS was Software as a Service. I believe it's going to be service as software." @generalcatalyst's Madhu Namburi on the AI roll-up thesis. Services is a $20 trillion market.... multiple times the size of software. not the PE playbook of debt and cost-cutting. Venture is buying legacy companies and using AI to drive growth
English
Using structured documentation like Mintlify as a source of truth for AI agents rather than humans is a game changer for maintaining system architecture context.
pro.statechange.ai/c/weekly-newsl…
English
Ray Deck @ statechange.ai retweetledi
Software used to be something you bought, installed, and kept. Now you describe what you want and throw it away when you're done. This changes everything about how we build companies.
English
Ray Deck @ statechange.ai retweetledi
YouTube
English
When did we forget in America that the institution that controls the guns controls the power (ie: the government). When did we get convinced that companies are more powerful than government rather than the idea that companies are limited in every way by the rules the government establishes and has the freedom to change at anytime. You know that government has more power cause that is who you want to use to regulate business. If government was less powerful than business it couldn’t regulate business. I’m not trying to say all businesses are good I’m just saying that from a power perspective this is pretty clear.
English
I call this the Tony Bennett principle. It's not about where you are physically. It's about where your heart and your mindset are. Think like Silicon Valley, and you can get the advantages too. Easier said than done, but doable.
Blake Robbins@blakeir
I tend to agree with this, then I remember @steipete quietly built Openclaw in Vienna. There are outliers everywhere.
English
Ray Deck @ statechange.ai retweetledi
Agentic coding is a form of machine learning. Generated code is best treated as a blackbox artifact whose behavior and generalization should be managed via empirical evaluation, like with any ML model.
François Chollet@fchollet
Sufficiently advanced agentic coding is essentially machine learning: the engineer sets up the optimization goal as well as some constraints on the search space (the spec and its tests), then an optimization process (coding agents) iterates until the goal is reached. The result is a blackbox model (the generated codebase): an artifact that performs the task, that you deploy without ever inspecting its internal logic, just as we ignore individual weights in a neural network. This implies that all classic issues encountered in ML will soon become problems for agentic coding: overfitting to the spec, Clever Hans shortcuts that don't generalize outside the tests, data leakage, concept drift, etc. I would also ask: what will be the Keras of agentic coding? What will be the optimal set of high-level abstractions that allow humans to steer codebase 'training' with minimal cognitive overhead?
English
How do you tell someone that everything they are building and publishing is AI slop?
English
Ray Deck @ statechange.ai retweetledi
cognitive expansion > cognitive offloading
so far, I see far more excitement about the latter, but the former requires creativity and is the real alpha
English
Mostly right, but while truth is often unkind, its delivery doesn’t have to be. A spoon full of sugar and all that. I always liked the framing of delivering hard counsel as to a “friend who owes your money.”
Brian Halligan@bhalligan
I don't remember where I found this, but its spot on.
English
Ray Deck @ statechange.ai retweetledi
Never underestimate how much time and effort you can waste by trying to automate a process you do not understand manually.
English
Ray Deck @ statechange.ai retweetledi
