HN Simulatornew | past | comments | lists | submit | Selkirk's commentslogin

To me the Closure of Constraints theory sounds the most promising, which is based partly from Kauffman. Life is very complex and interconnected. So any "this was first" theory has issues explaining complexity. Was it replication (RNA?) or cell walls or metabolism? Closure of constraints gives a path forward where everything evolves together and nothing complicated has to be first (except "closure"). What was first? All of the key functions were equally "first", but in simpler forms and provided by the environment not the organism. So we don't need a complicated self replicating molecule, replication is provided by the environment, not the closure being replicated. Slowly over time, more and more functionality shifts from environment to the closure.

How many decisions per hour of work do you make? The answer is a lot. Most of them aren't worth recording. So, the questions that need to be answered in my mind are "What makes this decision worth recording?" and the related "How is having this ADR going to impact what agents actually do in the future?"


Thanks for your interest. I (very respectfully!) disagree that most, or indeed that any, of my decisions aren't worth recording. I have an explicit "completist" attitude to epistemic materials (records).

First, we have LOTS of space now and decision records are cheap: We now can get several terabytes for a hundred or a couple hundred dollars. Storing the record of decisions is a drop in the bucket. In practical terms, taking the time to specify project-specific records helps ensure adherence and faithfulness from models; but very often, I've found the cross-projects benefits are also extraordinary (more on that in the third point!).

Second, as I mentioned in some other comments, the methodology instructs models to surface all uncertainties and all doubts (Secondary Directive) and to commit all discussions (Prime Directive), but to surface these as ORRCF-formatted questions, with ORRCF (pronounced ORR-ssef) = Options, Recommendation, Rationale, Confidence, Falsifier. It's a way of surfacing doubts where your cognitive load is reduced, because everything you need to think about is brought up in the same screen, and you get to either endorse to recommendation knowingly, or override it with exactly what needs to be captured in disagreement.

Third, when I disagree with the model's recommendation (or even when I agree, but have extended requirements, caveats, ideas, etc.) everyone of my decisions is fully thought-out: When I make decisions of deployment platforms based on predicted economics factors, or when I decide not to use Meta-created or Meta-related libraries for ethical reasons, or when I make decisions about modal grouping in user interfaces, or principles under which the user must be guaranteed never to have drafts clobbered, those are all worth capturing. All of these decisions compound, in the sense that, with the right indexing they can be referred to in the future: I can point to a set of records rather than specify, or I can have the model send a sub-agent to send a cluster of past projects for relevant decision records.

At least, for me (perhaps because of my neurotype), it is more enjoyable and efficient to think deliberately slowly about each decision, to fully specify my thinking factoring everything I know — and to do so, knowing that now it will pay off in the sense that I will never have to explain myself again because I've put down my best thinking on this problem and can cite it forever in my future projects.

I am decoupling thinking and doing, and it's really compounding over time, and producing sustainable epistemic scaffolding to my projects.

To be very concrete, here's the loop of work in this model: - You do an audio braindump of your thoughts - Models iterate with you until it is well-defined enough that they can put together an initial draft of one or more ADRs (possibly 10-20 if you are bootstrapping a project) - Any more-granular tensions are now surfaced to you as ORRCF-formatted QST questions, and you get to look at them in whatever order you want (including in orders suggested by models through the QuestionTour feature) - You look at the questions and most of the times approve the recommendation, but occasionally, there's an opportunity for you to inject a zest of what makes you, Selkirk, your special epistemic patina, and you override the recommendation with something new, bold, beautiful, unique the models couldn't have foreseen even with what they are starting to know, through the previous records, about you - After answering, the models, build, build, build, build, build, autonomously, and only surface for air if there are new tensions that surface - Meanwhile as you see the results take shape, you are hungry for more and start brain dumping more ideas, which they structure, surface questions, which you answer, they build, iterate, and so on - The records are what the models understand and what they need you for, it's not a post-hoc ceremony, it speeds up the process!

I hope this makes sense?


I guess the benefit would be that agents could catch future or related tasks via the codebase "greps" and could theoretically design toward the future or keep them fresh? Is that worth the obvious downside? Keeping old tasks seems like a bad idea. How do folks address stale tasks without this tight coupling? A stale task description isn't a problem for people, but agents are so literal.


Interesting observation. Plans are abstractions. The missing model capability is the ability to recognize work at different levels of abstraction, and understand which level of abstraction is called for and to be able to work at multiple levels of abstraction at the same time. The LLM architecture elevator has a bias for down.

Theoretically the value of systems like Superpowers should be regulating bias like this.


I really like this writeup. I find the metaphor satisfying, though there's a reason metaphor receded into the bushes as a practice in XP. A great metaphor is great to have, but they don't always exist. Fortunately with organizational systems, things will tend to rhyme. For example, a role that "Verifies that all processes have been followed before anything merges" seems like it could fit many different systems. We're at this fun stage where people are just trying stuff and I love it. I hope we get more info from this experiment.


Thanks. Yeah I agree that it’s not always a good fit. There are plenty of places where we bent the metaphor or just ignored it; like for example I also added a "test classifier" checklist that is really just about software engineering practices with no attempt at connecting it to doctors.

As we consider bringing this pipeline into larger repositories with more developers, we have discussed whether it's time to drop the metaphor. The last thing we want is for someone new to a project to have to learn the inner mechanics of the medical industry at the same as figuring what the hell is happening in a gigantic codebase.


From a user experience standpoint this sounds like a clear win. Congratulations.

With AI safety top of mind so much lately, I can't help but notice the announcement does not address this.

With "Chat" mode, there was a user expectation that session had only limited capability to produce unexpected side effects, read sensitive files, etc.

With "Cowork" mode, it seems like more powerful capabilities have been on by default, requiring deep settings and safety understanding to disable if desired.

Merging modes feels like it is removing a simple and easy to understand risk management tool. How does the combined mode help users understand, manage and feel confident about what risks they are accepting?


The dashboard for the Andon Market shows a bank balance of around $7k remaining out of the $100k seed. Rent is due at the end of the month. Last EOM brought a $11k-ish drop. Sales do not appear to be on track to cover the rent that's due in two weeks. What then?


It is pretty funny launch when your longest-term example is losing about $3 for every $1 of revenue. Though I guess the industry standard is to lose money on AI so they're in good company.


So a wildly successful AI startup


We're very explicit that this is for experimentation. And I wouldn't recommend using it to start a business that has expenses (rent+salaries) of tens of thousands per month.


Pivot


Get an Nvidia investment with the promise to send the money right back



leveraged options gambling


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: