BusinessNotes/wiki/sources/2026-07-06-sebastian-interview-ai-and-software-engineering.md

Sebastian Interview — AI and Software Engineering

#source #ai #interview #positioning

Source Metadata

• Date: 2026-07-06 (file date; the conversation itself is undated — described as a pre-webinar talk)
• Raw path: raw/sources/sebastian interview - conclusions and insights.md
• Source type: ~56-min recorded conversation, conclusions document. Built from a Whisper large-v3 auto-transcript, speaker-diarized; quotes lightly cleaned. A few rapid-crosstalk moments may be mis-attributed.
• Ingestion date: 2026-07-17
• Participants: sebastian (founder of virtido, ~11-yr software outsourcing firm) and eugene (interviewer; CV/embedded developer, content creator, builder of a custom Claude Code harness).
• Format note: A conclusions/insights write-up, one interpretive layer above the transcript. Distinct in origin from the other sources in this vault (which are the "15-06-2026 Inbox" batch) — this is an independent stream.

Core Claims

1. The debate is settled: AI can write software. The whole game is now how you use it. Cost of writing code trends to zero, so leverage moves entirely to directing and verifying the AI. See future-of-engineering-work.
2. Teams collapse from ~8 to 2–3. Scrum master + PM + requirements engineer + big dev team → one coordination/ownership role plus one or two agent-wranglers.
3. Bring-your-own-harness maximizes individual leverage but dies in the enterprise. Compliance and liability forbid ad-hoc per-developer setups → the business opportunity is compliant, company-standard managed harnesses for large regulated teams. See future-of-engineering-work, niche-selection.
4. Pure coding skill is commoditized; judgment, ownership, and relationships appreciate. Claude "levels" a 20-year veteran and a fresh grad to similar output.
5. Seniors became more valuable; juniors "completely irrelevant." A senior's real product is risk reduction. See seniority-and-ai.
6. Product ownership is the durable human skill — own the outcome a user sees, not the ticket. See product-ownership.
7. Relationships are the last non-commoditized asset, and only in-person builds them. Online outreach (agencies, cold call, email, LinkedIn, content, SEO) is "Big zero." See relationships-as-moat — this is the source's standout and most contrarian claim.
8. Decouple identity from profession, or "we will feel worthless in a couple of years."

Key Evidence / Details

Enterprise reality (Sebastian's lens): biggest clients forbid personal laptops — engineers use a centrally managed VM with zero ability to install their own tools. Reference points: a Roche SAP transformation ran ~1,200 engineers for years; banks first banned AI outright, now adopt cautiously "because it's just so good."

Why seniors win / the junior security argument: 20 years = knowing where things typically go wrong, so you don't let the AI make those mistakes ("AI does the same mistakes humans do because it's trained on our mistakes"). The junior risk is a security argument: reflexively clicking "allow for all future" is how "API keys are leaked, databases get dumped." A junior can't evaluate a 250-line bash script; a senior at least could. "Give a junior fresh out of university access to this almighty Claude and then access to the codebase — they will [wreck] it in two days."

The profile-picture story (ownership): an engineer implemented "change your photo," ticked every acceptance criterion, shipped it ugly (visible in the corner) because they never looked at the result. Reframe: stop thinking "what needs to be done" (tickets); think "what problem needs to be solved." "No one ever needed a programmer… people have problems that you are solving."

Planning / the printer anecdote: a colleague (Daniel) floated returning to waterfall; Sebastian disagrees — coordination overhead now exceeds the work. On a 2-person project (~1 day/week each) he is faster alone. A new printer needed a direct-printing protocol in Java (never written) working in 2 days; he went on-site and Claude Code solved it in ~30 min. His edge was knowing how to instruct and verify, not to code.

Connections — the mechanics: a connection forms not at first meeting but when you meet the same person in different circumstances → recognition → trust → referrals. Be memorable in your humanness (renovating a house, two kids, a cat) — not "I run a software company," which everyone forgets, because everyone now uses the same AI tools and everything online looks identical. Virtido's playful humans.verti.com "human badge" riffs on this. What works: showing up in person 2–4 days/week. What doesn't: "sales agencies, cold calling, email marketing, LinkedIn campaigns, content, SEO. Zero. Big zero."

Side points: open source will grow (code becomes nearly free to give away; Eugene's cynical read: OSS is largely marketing). Legacy/hobby niches persist (COBOL in banks — no training data; people who code "like driving an old-timer car") but not where time/quality/money matter. Outsourcing economics: great engineers exist everywhere, but a low-cost expectation selects for bad code — pay $10k/mo for excellent people; an $800/mo hire won't beat a good, well-paid engineer. AI helps non-programmers too: profile ~2,000 LinkedIn/HR contacts into Obsidian, build RAG over your own information — "you can put your whole life into a RAG."

Internal disagreements: personal vs company-managed harness (Eugene vs Sebastian); planning/waterfall (Daniel vs Sebastian); OSS motivation (Eugene: marketing; Sebastian: expects more); networking channels (Eugene invested in LinkedIn/articles; Sebastian: online outreach is a waste vs in-person).

Connections

• relationships-as-moat — the connections thesis, primary source
• product-ownership — ownership as the durable skill
• seniority-and-ai — senior/junior value inversion; risk reduction
• future-of-engineering-work — team collapse, harness, enterprise lockdown, identity, OSS, legacy
• client-acquisition-channels — Sebastian's "in-person only" is the sharp counter-pole to the outbound channel taxonomy
• methodology-as-moat — competing account of what the moat is (method vs relationship)
• ai-market-shift — the same AI shift, seen from the labor/enterprise side
• sebastian, eugene, virtido — entities

Open Questions

• How does AI transform huge (~1,200-engineer, multi-year) enterprise programs? Sebastian is candid he doesn't know.
• How do you unify wildly different personal workflows into one company process? (Eugene + a collaborator spent a month, failed.)
• Concretely, how does an individual build a connections network from a standing start? Principles given; the step-by-step is open.
• Sebastian's "online outreach = Big zero" is one founder's experience in an enterprise-services context. Does it generalize, or is it audience-specific (big enterprise buyers vs. SMB/startup buyers, where the other sources' channels may work)?
• Is eugene the owner of this vault? Internal details (the insin project, a custom Claude Code harness) match the vault owner's tooling. Recorded as inference, not fact.

Change Impact on Wiki

Second-batch ingest; the only source in this batch not from the "15-06-2026 Inbox" corpus.

• Created 4 new concepts primarily from this source: relationships-as-moat, product-ownership, seniority-and-ai, future-of-engineering-work.
• Created entities sebastian, eugene, virtido.
• Introduced the vault's sharpest live contradiction: in-person-only (Sebastian) vs. multi-channel outbound incl. online (AB Analytics / Tony / Rodenko) — recorded in client-acquisition-channels, sales-discipline, and overview.
• Introduced a second moat account (relationships) alongside the existing one (methodology) — tension recorded in relationships-as-moat and methodology-as-moat.
• Broadened the vault's domain in overview from "selling services in the AI era" to also include positioning engineering work/careers.