12) Greger Teigre Wedel: Når sjefen har svaret, tier resten

Truls Jørgensen og Audun Fauchald Strand — Heis Podcast · podcast · 2026-03-01 source ↗ Reader 🖥️

practical · mixed — Conversational interview that is mostly practical (how to lead, how to scale a product org, how to adopt AI) but threaded with theoretical observations about expertise, authority, and organizational dynamics. The title proposition — 'når sjefen har svaret, tier resten' — is a prescriptive claim about leader behavior, not a descriptive theory.

Structure

Key terms

• å være leder vs. å drive med ledelse Informal Leadership
  Wedel's distinction: being a 'leder' is the formal role with personnel responsibility; 'drive med ledelse' is the activity of moving a group toward an outcome, which one can do without title or authority. He values the second and is ambivalent about the first.
• når sjefen har svaret, tier resten Leader Restraint
  An operational claim: as soon as a leader signals they already know the answer (because they are technically hands-on, or because they speak too explicitly and detailed), alternative voices fall silent. 'Tier' carries the connotation of voluntary withdrawal, not being silenced by force.
• ta plass / la andre få plass Leader Restraint
  A learned leadership skill: taking space when you have passion and conviction, vs. consciously holding back to let team members occupy the conversation. Wedel reports being coached to 'tone down' early in his career and now sees it as a dynamic to manage in both directions.
• lederhatten / brette opp armene Hands-On Leadership & the Pendulum
  Two roles the hands-on technical leader can occupy: 'taking off the leader hat and rolling up the sleeves' to enter whiteboarding/maker mode. Wedel argues this works only if it is made explicit and contained — otherwise the leader role bleeds in and silences the team.
• produktorganisasjon
  A product org is one where the value sold is technology-driven and where engineers can deeply engage with an identifiable external customer. If you have to twist reality to name a customer, you don't have a product org — you have IT-supporting-business-processes, which requires a different operating model.
• first principles
  Wedel is mildly sceptical of the term as a buzzword but uses it to mean: the underlying logic that explains why five different methods all work — without which adopting any of them becomes 'theater' and 'incantations'.
• horisontal (alle må være mer horisontale)
  In the AI section: with AI as a tool, everyone must broaden across roles — designers commit code, backend engineers prototype frontends, PMs make prototypes — without claiming equal mastery. AI raises the floor on adjacent skills.
• verktøykassa / repertoar
  Wedel describes leadership growth as enlarging a toolbox of postures (coach, decider, hands-on, whiteboarder, director) and getting better at choosing which one to use in which context. Not better formulas; more options.
• guardrails (DORA-style) Guardrails
  Pre-existing engineering disciplines — tests, continuous delivery, code review — which AI does not replace but in fact requires more of, because AI amplifies whatever practice is in place (good or bad).

Propositions

• explicitcorroboratedWhen a leader displays that they already have the answer, team members go quiet rather than offer alternative views.conf: high
  extends ←AI Is Undermining Leaders’ Judgment. Here’s What to Do About It. · p8 — Wedel's 'when the boss already has the answer, other voices go quiet' mechanism, reapplied to AI: an algorithm's veneer of objectivity produces the same chilling effect on dissent that a visibly answer-holding human boss produces.
  supports ←Why Great Leaders Need More Than Good Judgment · p4 — Goldsmith's 'if you talk, now it's your idea, they salute the flag' is the same mechanism as 'when a leader displays they already have the answer, team members go quiet' -- both describe how a leader's own contribution silences others' ownership.
• explicitBeing technically hands-on is both an asset (credibility, judgment) and a liability (you carry answers into rooms that should be exploring).conf: high
• explicit'Being a leader' (the role) and 'doing leadership' (the activity) are distinct; one can do the second without ever holding the first.conf: high
• explicitLeadership development is not finding the right formula but expanding a repertoire of postures and getting better at picking the right one for the moment.conf: high
• explicitdevelopedInspiration about the vision is the leader's job; inspiration about how to reach it comes better from the team.conf: high
  extends ←The Power of Strategic Centering · p6 — McGrath's finding that centered-org leaders 'provide direction without dictating methods' converges, from a strategy-leadership context, on the heis podcast's claim that inspiration about the vision is the leader's job while inspiration about how to reach it comes better from the team — the same direction/method distinction surfacing independently in engineering leadership.
• explicitWithout personnel authority (e.g. as a principal engineer with 40 architects to orchestrate), one is forced to develop a non-coercive leadership style that transfers well into formal roles later.conf: high
• explicitAcquiring formal authority makes leaders 'blind to being wrong'; lacking authority disciplines you against that blindness.conf: high
• explicitMode-switching needs to be explicit ('now we're in whiteboarding mode, leader hat off') or it fails — the leader role leaks back in and silences others.conf: high
• explicitSometimes the discipline of holding back becomes its own failure mode: over-coaching when a decisive cut is what the team actually needs.conf: high
• explicitAn organisation is only a product organisation if it can identify a real external customer and let engineers engage with that customer in depth; otherwise it is IT-supporting-business-processes.conf: high
• explicitMost organisations fail to genuinely connect builders with users; the connection survives only 'in pockets', driven by specific people, not by structure.conf: high
• explicitAdopting a framework without understanding its underlying principle reduces practice to 'theater' and 'incantations'.conf: high
• explicitcorroboratedAI-assisted development is less a productivity revolution than a horizontal levelling: everyone must broaden, AI lifts the floor on adjacent skills, but the depth specialist still leads the depth.conf: high
  supports ←Netflix CPTO on AI and the future of product and tech roles | Elizabeth Stone · p9 — Corroborates 'horizontal levelling' with a concrete company example: everyone broadens, but Stone's rare deep-specialist niches are exactly the 'depth specialist still leads the depth' exception.
• explicitestablishedAI amplifies existing engineering practice: DORA-style guardrails (tests, CD, code review) become more important, not less, with AI-assisted development.conf: high
  supports ←We Studied 150 Developers Using AI (Here's What's Actually Changed...) · p6 — Both frame AI as amplifying existing engineering practice, making guardrails/discipline more important, not less, with AI assistance.
  supports ←TBM 428: Yes, Robot. Yes, Boss? · p1 — Independent restatement of AI-as-amplifier from the organizational (not just DORA-practices) angle.
  supports ←The Organization Is the Bottleneck · p10 — Independently asserts the distinctive claim that AI-accelerated shipping makes DORA-style guardrails (tests, pipelines, review) more important, not less.
• explicitWe do not yet have a working model for how juniors develop senior judgment when AI flattens the early production curve.conf: high
• explicitThe constant, not the technology, is learning — across decades, across language generations, across tool waves, the leader's job is to keep learning.conf: high
• explicitIt is irresponsible to lead in a domain you do not understand technically — many leaders 'sit and lead something they don't know anything about'.conf: high
• implicitcorroboratedNorwegian organisational culture (implied) makes the silence effect particularly strong because dissent is already costly to voice.conf: medium
  supports ←Strategi som regi i møte med teknologi: En samtale med Walter Qvam · p21 — Independent corroboration, from a different Norwegian leadership podcast, of the claim that Norwegian/Nordic organizational culture is particularly conflict-averse — though neither source actually tests this against a non-Nordic comparison (see this document's own gap g4).
• explicitPower, not just clarity, is what silences people: 'you are the boss too' is named as an additive force on top of merely being detailed.conf: high
• explicitThe hosts' shared experience: principal-engineer / director roles teach a non-coercive leadership style that does not transfer cleanly back to traditional command expectations.conf: high

Arguments

Narratives

• Cisco principal engineer: orchestrating 40+ architects without personnel authority. None reported to him; none would have accepted being commanded. He learned a non-coercive style that he now considers more durable than positional authority.
  → Supports the proposition that having authority makes leaders blind, and that the absence of authority is itself a developmental forcing function.
• Early-career coaching to 'tone down' his intensity in Cisco's multi-cultural, multi-timezone setting. He worked at being less intense, then later discovered he had over-corrected and needed to bring intensity back forward in specific situations.
  → Illustrates leadership-as-repertoire: there is no static answer to 'how intense should I be' — it varies by context, and the leader must keep adjusting.
• Flutter/Dart side project: built a 'first app' Flutter template during beta release, shipped it to both app stores, and it became a widely-used reference starter for the ecosystem (four-digit GitHub stars).
  → Demonstrates Wedel's claim that learning-for-its-own-sake is his motivation and his way of staying credible at the leadership level — but also shows what happens when a leader's side project becomes a public artifact.
• Cognite + Aker: a core success factor was that Aker portfolio companies opened their workplaces to Cognite engineers, giving rare access to chemical engineers, offshore operators, process engineers. This depth of access is what made software engineers actually able to model the user reality.
  → Supports the proposition that bridging the engineer-user gap is the hardest problem in industrial software and only happens 'in pockets' driven by structural commitment (here, ownership and access).
• Frustration of operational engineers (AKBP, Aker BP) at having developers 'in sandals' wander through asking questions while they are on tight project deadlines — and the analogy to NAV technologists trying to understand caseworker / jurist work for vedtak-systems.
  → Concrete instance of the structural difficulty in connecting builders with domain users — the time and culture asymmetries are not minor.
• Omny Security AI hackathon: Wedel set a deliberate challenge — designers should commit code, backend engineers should prototype frontend, PMs should make prototypes. Not because their output goes into production but because it elevates the inputs of those closest to the customer.
  → Supports the claim that AI levels horizontally: the value is not that everyone becomes equally skilled but that the team's collective inputs are sharper.
• Advice to his own children (21, 18, 11): be excellent at the basics (math, physics, chemistry, reading, reasoning, communication, group work), and be an experimenter with AI.
  → Personal data point on what he believes is durable vs. ephemeral — fundamentals plus experimentation, not specific technologies.
• Hosts (Truls / Audun) recall meeting Wedel for the first time and being struck by how he managed to remain technically engaged (coding) while leading at scale — they cite him as the reason they recall the role being possible.
  → Frames the conversation: the hosts treat Wedel as an instance of the rare hands-on technical leader, which is why the silence problem matters here.
• Jensen Huang / NVIDIA mental model: the organisation as a computer, with processors, sub-processors, GPUs each doing their piece. Wedel cites it approvingly.
  → Used to ground his bachelor-in-psychology-with-group-dynamics-focus view: roles in an organisation are like components in a system, and the leader's design job is to shape interactions rather than command.

Questions addressed

• ✅ What does an experienced technical leader actually do differently from a beginner?
  They hold a larger repertoire of postures and make sharper situational choices about which to deploy — including when to suppress their own answer.
• ✅ What makes a technically hands-on leader an asset vs. a liability to their team?
  The asset is judgment and credibility. The liability is the silencing effect when the leader's answer enters the room. The discipline is explicit mode-switching and selective withholding.
• ✅ What is a product organisation, really?
  One whose value is technology-driven AND that can identify and deeply engage an external customer. Anything else is IT-supporting-business-processes and needs a different operating model.
• ✅ How does AI-assisted development change engineering organisations?
  It levels horizontally and amplifies existing practice. DORA-style guardrails become more important. The first-order effect is broadening, not replacement.
• ⭕ How do juniors develop senior judgment when AI flattens the early production curve?
  Wedel explicitly names this as unsolved.
• ⭕ What identity does a senior engineer hold when AI absorbs much of their daily craft?
  Wedel names 'identity problem' as an emerging concern but does not resolve it.
• ⭕ Is the silence effect culturally Norwegian, or universal?
  Not addressed — Wedel treats the silencing effect as a working assumption.
• ✅ When IS having (and using) the answer the right move?
  Partially — Wedel names the over-coaching failure mode and the moment for cutting through ('det er tre gode alternativer her, men vi tar den'). The conditions are not fully specified.

Gaps

• incompleteCultural specificity is unexamined: Wedel never tests whether 'when the boss has the answer, the rest stay silent' depends on Norwegian (or Nordic, or hierarchical) defaults. In flatter or more confrontational cultures, the chilling effect may be weaker — or it may show up differently (silent disagreement vs. silent compliance).
  Read next: Cutler TBM 423 on undiscussables (which treats power-asymmetry-creating-silence as general, not culturally specific) and Edmondson on psychological safety across cultures. Also Hofstede / Trompenaars on power distance dimensions in Nordic vs. other workplaces.
• incompleteDecision criteria for posture-switching are underspecified. Wedel says leaders should choose among coaching, deciding, hands-on, whiteboarding, director postures situationally — but he names only one explicit trigger (the over-coaching failure mode that calls for cutting through). When to take which posture is left as 'judgment'.
  Read next: Kent Beck on tidy-first vs. mess-first decision rules; Heifetz on adaptive vs. technical work; Cutler's TBM 418 on campfires/trails/quests as posture-selection frame.
• incompleteThe interaction between explicit mode-switching ('lederhatten av') and team-level psychological safety is asserted but not unpacked. Saying 'we're in whiteboarding mode now' presumably only works if the team trusts that the leader will actually honour the signal — which is itself a function of trust built elsewhere.
  Read next: Edmondson on psychological safety as accumulated track-record; Schein on rituals and their preconditions; Pocock's 'alignment is not AFK' point about why alignment work must be live, not formal.
• incompleteThe junior-development problem is named but not theorised. Wedel says we don't know how to grow juniors when AI flattens the early production curve. He gestures at 'fundamentals matter more' but doesn't say how organisations should change apprenticeship.
  Read next: Recent writing on AI and the apprenticeship pipeline (Pragmatic Engineer FDE piece, Addy Osmani on orchestration tax, Garg on reducing friction in AI development).
  closed byHow to Turn Individual Talent into Organizational Excellence — Offers a theorized, systemic answer to 'how organisations should change apprenticeship': make senior-as-teacher a formal role expectation tied to status and promotion, engineer deliberate exposure to higher-stakes decisions, and build feedback into recurring work routines — rather than leaving apprenticeship personality-dependent. Only partially closes the gap, since it does not address the AI-specific mechanism (what replaces automated on-ramp tasks) that motivated the original question.
• incompleteAI-as-amplifier is asserted in both directions (good practice → better; bad practice → worse) but the magnitude of asymmetry is not addressed. If AI amplifies bad practice faster than good, the policy implication ('defer AI until guardrails exist') may be stronger than Wedel states.
  Read next: DORA 2025 detailed findings; the Dave Farley YouTube series Wedel mentions; any empirical work on AI adoption velocity vs. defect rates.
• incompleteThe six product-org first principles are alluded to but not enumerated in the conversation. Only the first (deep user/customer connection) is discussed.
  Read next: stuff.greger.io blog series on product organisation first principles.

Related documents

• The engineering manager role is splitting in two
  Hands-On Leadership & the Pendulum Informal Leadership
• Alex Ewerlöf's Post
  Hands-On Leadership & the Pendulum Informal Leadership
• How to Build a Superteam That Keeps Getting Better
  Hands-On Leadership & the Pendulum Leader Restraint
• 3) Christin Gorman leder uten tittel
  Hands-On Leadership & the Pendulum Informal Leadership
• 4) Morten Eikeland: Det finnes alltid et hierarki
  Hands-On Leadership & the Pendulum Informal Leadership
• Update on Software Engineering Career Ladder
  Hands-On Leadership & the Pendulum Informal Leadership