Everything is code, but the alpha is in the parts that do not scale with code. Power, fabs, materials, bandwidth, governance, and measurement. Agentic IDEs compress the cost of building toward zero, so value shifts upstream into bottlenecks and downstream into selection. The real question becomes: who owns the constraints, and who owns the scoring layer that decides what is useful.
This is exactly the signal I am looking for! love the comment!
You have perfectly articulated the 'Inversion Thesis' I’ve been wrestling with. When the marginal cost of code hits zero, the premium on physics (power/fabs) and judgment (governance/scoring) goes infinite. There's also no shortage or pricing signals from the public equity markets for this reality being shaped right now.
We are moving from an era of construction (who can build the app?) to an era of curation (who defines the truth functioning of the app?) and validators (how was sth built in the first place?). That 'scoring layer' you mentioned? That is the new regulatory capture.
I have so many ideas and conflicting thoughts here. I'll have to explore these in SIGNAL in the future. Glad to have you thinking along with me.
Btw: I plan on digging deeper into 'Who owns the constraints?' in a future SILICON deep dive. Great addition to the discussion. Thx
Spot on. As build costs collapse, value shifts to what does not scale with code: physics constraints (power, fabs, packaging, bandwidth) and the judgment layer (provenance, benchmarks, auditability). I like the “scoring layer” framing as one of the main capture surfaces. When construction commoditizes, whoever controls evaluation, ranking, and verification quietly governs outcomes.
If you are open to it, I would love to do a short co-note or Q&A sometime. Your sand-to-server framing plus my scarcity/bottleneck map (who owns constraints, who owns scoring, where rents accrue) feels like a clean fit.
This means a lot coming from you. I do consider your comments very high signal.
I am definitely open to collaborating down the road, but I want to be transparent: I am still in the early 'bootstrapping' phase of this publication. I’m currently focused on stabilizing my writing cadence and refining the voice before I open up the API for external collaborations.
I want to make sure that when we do something, I can bring the same level of depth to the table that you do!
Let’s keep the channel open? I’d love to circle back once I’ve got a few more deep dives under my belt.
Everything is code, but the alpha is in the parts that do not scale with code. Power, fabs, materials, bandwidth, governance, and measurement. Agentic IDEs compress the cost of building toward zero, so value shifts upstream into bottlenecks and downstream into selection. The real question becomes: who owns the constraints, and who owns the scoring layer that decides what is useful.
This is exactly the signal I am looking for! love the comment!
You have perfectly articulated the 'Inversion Thesis' I’ve been wrestling with. When the marginal cost of code hits zero, the premium on physics (power/fabs) and judgment (governance/scoring) goes infinite. There's also no shortage or pricing signals from the public equity markets for this reality being shaped right now.
We are moving from an era of construction (who can build the app?) to an era of curation (who defines the truth functioning of the app?) and validators (how was sth built in the first place?). That 'scoring layer' you mentioned? That is the new regulatory capture.
I have so many ideas and conflicting thoughts here. I'll have to explore these in SIGNAL in the future. Glad to have you thinking along with me.
Btw: I plan on digging deeper into 'Who owns the constraints?' in a future SILICON deep dive. Great addition to the discussion. Thx
Spot on. As build costs collapse, value shifts to what does not scale with code: physics constraints (power, fabs, packaging, bandwidth) and the judgment layer (provenance, benchmarks, auditability). I like the “scoring layer” framing as one of the main capture surfaces. When construction commoditizes, whoever controls evaluation, ranking, and verification quietly governs outcomes.
If you are open to it, I would love to do a short co-note or Q&A sometime. Your sand-to-server framing plus my scarcity/bottleneck map (who owns constraints, who owns scoring, where rents accrue) feels like a clean fit.
This means a lot coming from you. I do consider your comments very high signal.
I am definitely open to collaborating down the road, but I want to be transparent: I am still in the early 'bootstrapping' phase of this publication. I’m currently focused on stabilizing my writing cadence and refining the voice before I open up the API for external collaborations.
I want to make sure that when we do something, I can bring the same level of depth to the table that you do!
Let’s keep the channel open? I’d love to circle back once I’ve got a few more deep dives under my belt.
Thank you for the kind words, I really appreciate it. And honestly, you are already bringing depth.
My Substack is not even 6 months old either, so I am also in startup mode and learning every day, but I have the resolve to keep sharpening the craft.
Absolutely, let’s keep the channel open.