Agentic Engineering: When Supervising Is Worth More Than Typing

STRATEGY · TECHNOLOGY

Agentic Engineering: When Supervising Is Worth More Than Typing

The discipline that follows "vibe coding" promises speed without sacrificing quality. But it only works in organizations that invest as much in verification as they do in models.

In February 2025, Andrej Karpathy, a co-founder of OpenAI and former head of AI at Tesla, popularized the term "vibe coding": giving in to the vibes, asking a model for code, and forgetting that the code exists. A year later, Karpathy himself declared his coinage obsolete. Models had improved enough that professional teams stopped typing most of their software and began directing agents that write it. For this more disciplined and supervised use, he proposed another name: agentic engineering.

The label matters because it changes the executive's question. It is no longer whether to let AI write code, but what system of controls makes what it writes trustworthy. This article defines the concept, sets out its objectives, describes the means of developing it, reviews documented cases, and assesses its prospects and its limits.

What it is, and what it is not

Karpathy justified the name with two words. "Agentic," because the new default is that you do not write the code directly most of the time: you orchestrate agents that do, and you provide oversight. "Engineering," because there is art, science and expertise in it, and because it is something one can learn and get better at.

The difference from vibe coding is one of rigor. The latter raises the floor: almost anyone can build prototypes, scripts and experiments. Agentic engineering raises the ceiling: it uses agents to produce software you can trust, held to the usual standards of security, reliability, architecture and maintainability.

Three traits set it apart. First, intentional delegation: the human defines the goal, constraints and acceptance criteria before releasing the agent. Second, verification: nothing reaches production without tests and review. Third, accountability: the engineer remains answerable for what is deployed, even without having written a single line.

In practice, the engineer's work shifts: less code and more specifications, plans and acceptance criteria; more reviewing and less typing. The skill that earns a premium is no longer typing speed but the ability to break a problem down, anticipate how it can fail, and confirm that it was actually solved.

A word of candor: it is not a formal standard or a certification. It is a label the industry adopted quickly and that each organization interprets in its own way, which makes it necessary to define what you mean by it before investing.

What it is for: the objectives

The first is speed without debt: multiplying output without accumulating hidden flaws that must be paid for later. The second is verifiable quality: turning trust into a measurable process instead of a hunch. The third is redeploying talent: moving people from typing toward design, architecture and judgment, which is where an agent still needs direction.

The fourth objective is scaling beyond the individual: moving from the personal assistant to fleets of agents working in parallel on well-bounded tasks. The fifth is governing risk: leaving a trail of what each agent did, with what permissions and under whose approval.

How it is developed: the means

The experience of pioneering companies points to six levers, none of which is the model itself.

Context. An agent performs according to what it knows about the project. Clear specifications, up-to-date documentation and explicit repository conventions are worth more than a clever prompt. Making systems legible to machines becomes an engineering investment.

Isolated, reproducible environments. Agents need somewhere to work without risk: disposable environments with the same tools people use. An analysis by the firm Ona of Stripe, Ramp and Spotify argues that the obstacle was not model quality but the development environment.

Automated verification. Tests, continuous integration, static analysis and security scanning act as a safety net: the agent runs, fails, fixes and iterates before a human looks at the result.

Least privilege and traceability. Curated tools, restricted access and a log of every call. The OWASP project has already cataloged risks specific to agentic applications, such as goal hijacking and tool misuse, which two years ago did not exist at enterprise scale.

Focused human review. The person reviews the plan and the outcome at defined checkpoints rather than reading every line. There is a trap: as volume grows, review can degenerate into a formality. It is a legitimate doubt that should be measured, not assumed resolved.

Training and cultural change. According to a study by Microsoft researchers on the rollout of command-line agents, early use spread mainly among colleagues rather than by mandate. The gains do not distribute themselves: they require training and a redesign of the workflow.

Success stories

Stripe. Its "Minions" are unattended agents: they start from a Slack message or a ticket and return a change ready for review. According to the company itself, in February 2026 more than 1,300 pull requests per week were merged without a single line written by humans, all reviewed by people. Stripe built on an open-source base rather than adopting commercial tools, precisely because of the requirement to operate without continuous supervision.

Rakuten. In its January 2026 trends report, Anthropic credits Rakuten with cutting feature delivery time from 24 days to 5. Separately, a case study published by OpenAI in March 2026 reports that the company halved its mean time to resolve incidents with Codex. In both cases the figures are published by vendors.

TELUS. According to coverage of the same report, the Canadian telecom accumulated more than 500,000 hours saved with roughly 13,000 AI solutions built in-house.

Ramp. Its internal agent, Inspect, was already generating more than half of merged pull requests in February 2026, according to an Ona white paper based on engineering blogs.

The common pattern, an independent review of cases notes, is that each company scaled thanks to verification and not thanks to the agent alone.

What the numbers do not say

These results deserve caution. Faros AI telemetry on 22,000 developers, from April 2026, shows epics per developer up 66.2%, but also bugs per developer (54%) and incidents per pull request (242.7%). Speed and quality do not automatically move in the same direction.

The Microsoft study found about 24% more merged pull requests among adopters, but its authors caution that they measure an output indicator, not delivered value. And according to Anthropic's report, developers use AI in around 60% of their work but fully delegate only 0% to 20% of tasks. The trust gap, more than model capability, is today's brake.

Prospects

Karpathy argues that the leverage of top-tier agentic engineering is very high, though not perfect: it requires high-level direction, judgment and taste. Three trends deserve watching.

The first is growing autonomy: agents chaining more steps without human intervention, which raises the value of well-designed checkpoints. The second is expansion beyond engineering: non-technical teams building their own tools, as suggested by the 89% adoption Zapier reports across its organization. The third is a premium on judgment: when producing code costs less, deciding what to build and spotting what is wrong are worth more.

There are also second-order risks. If agents write almost everything, organizations must prevent erosion of their understanding of their own systems: a team that does not understand its code cannot supervise it. That is why continuous training, rotation in reviews and a record of design decisions are not a luxury but part of the control.

For regulated sectors, such as banking or auditing, the parallel with internal control is natural: segregation of duties (the one who generates does not approve), audit trails and access management, now applied to agents.

What leaders should do

Three questions chart the path. Which tasks are delegable? Start with bounded ones that have clear tests. How will we know the result is correct? Define acceptance criteria and tests before multiplying agents. Who is accountable? Assign a human owner to every change that reaches production.

Measure outcomes, not activity: defect rate, incident recovery time and delivered value, not just the amount of code generated.

The competitive advantage will not go to whoever has the most agents, but to whoever has the best supervision.

Glossary

Agent. An AI system that does not just answer but plans and carries out multi-step tasks using tools such as editors, terminals or APIs.

Agentic engineering. Disciplined, supervised use of AI agents to build software while keeping professional quality standards. The term was proposed by Karpathy in February 2026.

Acceptance criteria. Conditions a change must meet to be considered correct, defined before the agent starts work.

Checkpoint. A defined point in the workflow where a human reviews and approves the plan or the outcome.

Continuous integration (CI). An automated process that builds the code and runs tests on every change.

Context. The information an agent receives about the project (specifications, documentation, conventions) and that conditions the quality of its work.

Goal hijacking. An attack or failure in which an agent is diverted from its original objective, for example through malicious instructions hidden in the data it reads.

Guardrails. Technical and procedural limits (permissions, tests, approvals) that constrain what an agent can do.

Human in the loop. A design in which a person reviews or approves the agent's actions at key points.

MTTR (mean time to resolve). Average time taken to resolve an incident or restore a service.

One-shot (unattended) agent. An agent that takes a task from start to finish without intermediate human interaction, returning a result for review.

OWASP. An open community that publishes security guidance, including a list of risks specific to agentic applications.

Orchestration. Coordinating one or more agents, assigning tasks, resources and limits, and integrating their results.

Pull request (PR). A proposal to merge a code change, which others review before it enters the main codebase.

Sandbox. An isolated, disposable environment where an agent can run code without risk to production systems.

Static analysis. Automated inspection of code, without running it, to detect errors or vulnerabilities.

Telemetry. Automatic collection of usage and performance data, used here to measure the effects of adopting agents.

Vibe coding. Style of development popularized by Karpathy in February 2025, in which the user asks a model for code and accepts the result without reviewing it closely, suited to prototypes and experiments.

Sources consulted

–    Karpathy, A., on "agentic engineering" (Feb. 2026), per The New Stack, "Vibe coding is passé" (Feb. 10, 2026): thenewstack.io/vibe-coding-is-passe

–    Stripe, "Minions: Stripe's one-shot, end-to-end coding agents", Parts 1 and 2 (Feb. 2026): stripe.dev

–    Anthropic, 2026 Agentic Coding Trends Report (Jan. 21, 2026): resources.anthropic.com/2026-agentic-coding-trends-report

–    OpenAI, "Rakuten fixes issues twice as fast with Codex" (Mar. 11, 2026): openai.com/index/rakuten

–    Ona, white paper on background agents (2026): ona.com/files/background_agents_white_paper.pdf

–    Murphy-Hill, Butler and Savelieva, arXiv 2607.01418 (Jul. 2026), and Faros AI telemetry (Apr. 2026), both cited in developertoolkit.ai/en/strategy/success-stories

After the ISS: The Business of Living in Orbit

STRATEGY · SPACE INDUSTRY

After the ISS: The Business of Living in Orbit

The most expensive laboratory in history is nearing retirement. Replacing it is a business problem more than an engineering one, defined by three gaps: demand, access, and time.

Analysis · Data as of November 3, 2026

The idea in brief

The problem. The ISS will be deorbited in 2031, and the United States is betting that private stations will replace it. Yet in March 2026 NASA itself said it did not believe a commercial business case yet exists in low Earth orbit.

The thesis. The outcome will turn less on habitat design than on three gaps: demand beyond NASA, reliable crew access, and continuity over time.

The takeaway. Treat the agency as an anchor customer, not a sponsor; diversify customers and critical suppliers; demand outcomes, not process.

On June 5, 2026, five astronauts aboard the International Space Station were ordered to shelter in the docked Crew Dragon capsule and prepare for emergency departure. Roscosmos was about to attempt a repair on a persistent leak in its Zvezda module. It paused, NASA backed the decision, and the crew returned to normal operations. Nobody left. Yet the episode captured the underlying dilemma: critical, cooperative, aging infrastructure nearing its end, and a successor that does not yet exist.

Replacing the ISS is a management case as much as an engineering one. For the first time, a government is trying to move from owner-operator of a frontier asset to one customer among many. That shift worked for cargo and crew transport. It is far harder when what you are buying is a place to live.

What the ISS proved

Space stations began as instruments of prestige: the Soviet Salyut 1 in 1971, America’s Skylab in 1973, then Mir. The ISS changed the logic. Its first module launched in 1998 and continuous occupation began in November 2000. Five agencies, from the United States, Russia, Europe, Japan, and Canada, sustained for a quarter century an operation that has hosted hundreds of astronauts from more than twenty countries. Its greatest achievement is organizational: it showed that partners with diverging political interests can share a critical asset for decades.

It also revealed the cost of aging. Since 2019, the PrK transfer tunnel of the Zvezda module has leaked; engineers attribute this to microscopic cracks, and sealing attempts have not produced a definitive fix. Retirement is planned for 2031. In June 2024 NASA hired SpaceX, for $843 million, to build a Dragon-derived deorbit vehicle that will guide the station into an uninhabited stretch of ocean. NASA’s inspector general warned that the schedule, roughly five and a half years from award to launch, is shorter than the eight and a half years major agency programs typically take.

The anchor-customer bet

The U.S. strategy replicates the formula that worked for cargo and crew: NASA funds early development, certifies, and then buys services. In December 2021 it split $416 million among three teams: one led by Nanoracks ($160 million), one by Blue Origin ($130 million), and Northrop Grumman ($125.6 million). Five years later, the formal competition is only now beginning.

In March 2026 the agency proposed a radical shift: buy a core module attached to the ISS to which commercial modules would connect, saying it saw no commercial business case yet. Industry pushed back, arguing that operating while attached to a government station erodes the very benefits of a commercial platform. In June NASA backed down. On October 9 it issued the final request for proposals: bids are due December 8, at least two companies will receive initial awards by April, and at least one will later advance to full development.

The managerial lesson is uncomfortable: the anchor customer changed the program’s design more than once before committing. For anyone investing their own capital, that uncertainty is a cost no spreadsheet captures.

The contenders

The competition is open to any company, but four names dominate. Each embodies a different thesis about what wins such a race.

 

 

 

 

 

 

 

Vast CEO Max Haot calls it a race in which whoever reaches orbit first will be best positioned. Vast bets on speed, Axiom on operating history, Starlab on its partners’ industrial heritage, and Blue Origin on manufacturing scale. None of these advantages alone closes the three gaps that follow.

The three gaps

The demand gap. Today the only buyer with sustained budget is NASA. Companies argue that sovereign astronauts, meaning countries without their own program paying to fly a citizen, along with private research and microgravity manufacturing, can form a market. The agency does not consider this proven. Until a second heavyweight customer exists, a commercial station is in practice a government contractor, with the concentration risk that implies.

The access gap. The paradox is that a station could be ready and nobody could reach it. SpaceX has privately indicated that Crew Dragon will likely not be available for these stations as it transitions to Starship. Axiom CEO Jonathan Cirtain said he has no pivot plan, because his training program depends on SpaceX vehicles. NASA responded with a $359 million contract modification for Boeing’s Starliner, but Administrator Jared Isaacman ruled out a new government-funded commercial crew program. The final request offers NASA-provided transportation only for the first four missions and tells companies to assume none afterward.

The time gap. The ISS ends in 2031. Haven-1 aims to fly in early 2027, after slipping from mid-2026, but it is a demonstration module, not a NASA-qualified station. Starlab targets 2029. In an industry where schedules slip routinely, any overlap between the ISS and its successor is fragile. The strategic risk is that the United States and its partners lose continuous human presence in low Earth orbit while China maintains its own.

The other board

China completed its three-module Tiangong station in late 2022 and runs it with three-person crews rotating every six months. It has announced a six-module expansion, is preparing its first one-year stay, and expects a Pakistani astronaut as its first foreign visitor. In late 2025, when a crew was stranded by damage to its spacecraft, it launched a replacement capsule uncrewed, a display of responsiveness few operators can match. India says its Bharatiya Antariksh Station will launch a first module in 2028 and be completed around 2035, a schedule not yet tested. Russia is studying its own station, with an uncertain timeline. The emerging map is less a successor than an archipelago: state stations from China, India, and perhaps Russia, and Western commercial platforms competing for the same sovereign clients.

What leaders should learn

Beyond aerospace, the transition offers four lessons for any executive who depends on a dominant customer or an infrastructure that is ending.

1.    An anchor customer is not a business model. It is a starting condition. The companies that thrive will build a portfolio of buyers beyond NASA from day one.

2.    Avoid a single supplier at the critical link. Crew transport is the sector’s chokepoint, and nearly every station depends on one provider. Map where that link sits in your own chain.

3.    Sequence your options. Vast intends to prove Haven-1 first and scale with Haven-2 afterward. Testing before committing large-scale capital lowers the cost of being wrong.

4.   Demand outcomes, not process. Cirtain said that for Axiom’s lunar spacesuit design review the company had to deliver 30,000 documents and close 4,000 issues. Applied to a station, he warned, you would never reach the finish line. NASA says it trimmed requirements in the final version but did not quantify the cut.

Conclusion: the handoff is won before launch

The ISS will be remembered as history’s greatest feat of technical cooperation, and its end is a risk-management decision more than a spectacle. The open question is not whether private stations will exist, since several are being built, but how many will survive without a single buyer, who will carry crews to them, and whether they will arrive in time to keep human presence unbroken.

The next dates give a clear signal: proposals are due December 8 and first awards are expected by April 2027. By then we will know who remains in the race and who must regroup or change markets. History suggests the winner will be whoever first closes the gap that looks least like engineering: finding someone else to pay to live up there.

SOURCES

Foust, J. “The commercial space station competition finally begins.” The Space Review, October 19, 2026.

Payload. “Payload Field Guide: Commercial LEO Destination.” Updated June 3, 2026.

Axios. “Why the International Space Station keeps leaking.” June 8, 2026.

SpaceNews. Coverage of the NASA inspector general audit of the ISS and the deorbit vehicle.

Payload. “Vast Delays Haven-1 Launch to 2027.” January 20, 2026.

SpaceNews. “China set for its first one-year human spaceflight mission.” February 27, 2026.

NASA. U.S. Deorbit Vehicle selection announcement, June 2024.

Note: company launch dates are stated targets and have changed often. Verify figures before citing them in publication.

Glossary

Anchor customer: A large, reliable buyer whose commitment makes a new market viable and gives other customers confidence.

CLD (Commercial LEO Destinations): NASA program to fund and certify privately owned stations in low Earth orbit, buying services from them instead of owning them.

Crew Dragon: SpaceX’s crewed capsule, currently the main U.S. vehicle for carrying astronauts to the ISS.

Deorbit: Deliberately lowering a spacecraft so it re-enters the atmosphere and burns up, ideally over open ocean.

GFTSS: Government Furnished Transportation and Support Services: NASA-provided crew and cargo transport offered in the final CLD solicitation.

Haven-1: Vast’s single-module demonstration station, planned to host four-person crews on short stays.

LEO (low Earth orbit): The region of space up to roughly 2,000 km altitude, where the ISS and Tiangong operate.

Roscosmos: The Russian space agency, operator of the ISS’s Russian segment.

RFP (request for proposals): The formal document in which a buyer invites companies to bid for a contract.

Safe haven: An emergency configuration in which the crew shelters in a docked vehicle, ready to leave the station.

Sovereign astronaut: An astronaut flown on behalf of a national government, often by a country without its own crewed program.

Starliner: Boeing’s crew capsule, still being certified after development delays.

Starship: SpaceX’s next-generation super-heavy launch system, intended to replace its current rockets over time.

Tiangong: China’s three-module space station, continuously crewed since its completion in late 2022.

USDV: U.S. Deorbit Vehicle: the SpaceX-built spacecraft that will push the ISS out of orbit at the end of its life.

Design as a System: How Apple Keeps Its Edge Without Jony Ive

Analysis · October 2026

The idea in brief

•         Apple's design is not a matter of aesthetics but an organizational capability that integrates hardware, software, and silicon. That is what makes it hard to copy.

•         Jony Ive's departure in 2019 exposed the risk of depending on one person: according to several reports, the studio lost influence and nearly all of its Ive-era talent.

•         The current answer is an unusual governance bet: CEO John Ternus has, in practice, taken over the direction of design. It turns a personal virtue into a discipline of top management, but it moves the key-person risk rather than removing it.

 When Steve Jobs returned to Apple in 1997, the company was on the brink of bankruptcy. What saved it was not exclusive technology but an idea: that a translucent, colorful computer, the 1998 iMac, could be desirable. Nearly three decades later, Apple became the first company to surpass $3 trillion in market capitalization, and its new CEO summed up the bet for employees in a simple line: Apple is the most beautifully designed thing most of its customers own, and it must stay that way.

The uncomfortable question is whether that advantage is durable. Jony Ive, the designer who shaped the iMac, the iPod, the iPhone, and the iPad, left the company in 2019. Most of his longtime collaborators have since left as well, and for seven years no one filled the central role he held. This article makes three arguments: that Apple's design is a competitive advantage precisely because it is a system and not a style; that the system weakened after Ive's exit; and that Apple is trying to rebuild it with a new governance model that has clear virtues and risks worth naming.

What the advantage really is

It is tempting to explain Apple by the beauty of its products. That is an incomplete reading. The advantage comes from three mechanisms that design makes possible.

The first is integration. Apple designs the device, the operating system and, since the 2020 move to Apple Silicon, the processor. That chain allows decisions a fragmented competitor can hardly make: giving up a millimeter of thickness to gain battery life, or tuning a software animation to the exact characteristics of a display. An Android phone maker depends on Google for the operating system and on third parties for the chip; a Windows PC maker shares control with Microsoft and the processor suppliers. A rival can match a camera. It is much harder to match the feeling that everything works as a single object.

The second is the ecosystem. Earbuds that pair instantly, a watch that unlocks the computer, a message that hops from one device to another: each product is better because the others exist. That coherence raises the cost of switching brands. Here design is not the casing but the architecture of the experience.

The third is pricing power. A product perceived as the most carefully made on the market sustains a premium and margins that in turn fund more precision manufacturing and more development. It is a virtuous circle, but it depends on the perception of quality being renewed with every generation.

What makes this advantage hard to copy is not a technical secret but coordination: engineering, design, operations, and marketing aligned behind a single hierarchy of priorities. That is exactly the asset that dependence on one person puts at risk.

The Ive model and its fragility

Under Jobs and Ive, the design studio worked like a small, protected organism: few members, direct access to the CEO, endless prototypes, and an informal rule that design had a decisive voice over the product. The obsession with materials, the simplicity inherited from Dieter Rams, and the intolerance of careless detail were not ornaments. They were a shared decision criterion.

But that model rested on two personal relationships. In 2015 Ive stepped back from day-to-day management as chief design officer; in 2019 he left to found LoveFrom. Evans Hankey took over industrial design and the studio came under the supervision of Jeff Williams, then chief operating officer. According to Bloomberg reports picked up by several outlets, the studio carried less weight internally, some described it as a "service bureau," and nearly all of the Ive-era designers ended up leaving. Hankey also left, and Alan Dye, who led interface design, moved to Meta.

It is worth being precise: none of this means Apple stopped shipping admired products. Its own silicon, for example, kept improving. But the difference between maintaining a product line and maintaining a design culture is what separates a present advantage from a future one. Over time, according to those reports, operational efficiency, manufacturing scale, and business execution gained ground over design as a strategic voice.

How Apple sustains it today: four mechanisms

1. From genius to governance. Apple has not started a search for an outside design chief. Instead, John Ternus, who took over design oversight in late 2025 and has been CEO since September 1, 2026, has become the de facto head: according to Bloomberg and journalist Mark Gurman, he works in the studio several times a week, sometimes more than once a day. The change is fundamental. Under Cook, design answered to an intermediate executive; today it answers to the person who makes the final call. The upside is speed and an internal signal: the whole organization understands that design matters again. The risk is replacing dependence on Ive with dependence on the CEO, who must also handle component prices, artificial intelligence, and the supply chain.

2. Two studios, one standard. Industrial design is led by Molly Anderson; human interface design by Steve Lemay, who succeeded Dye. Apple has presented them together at recent launches, a deliberate gesture, because the seam between hardware and software is where its advantage lives. Ternus sits at that junction, and executives such as Craig Federighi and Greg Joswiak retain a voice in the company's overall aesthetic.

3. Engineering and design on the same side. Ternus comes from hardware product engineering. A recent analysis notes that under Ive, design and engineering often clashed, and that Ternus wants them to work together. It corrects a flaw of the earlier model: beauty that was sometimes paid for in manufacturing difficulty. And the ability to manufacture at scale what others cannot produce is part of Apple's moat, not a logistical detail.

4. Rebuilding the talent. Apple has listed Anderson as vice president of industrial design on its leadership page and has brought in new profiles, such as Belgian furniture designer Julien Renault. It is fresh blood, not clones of the Ive era. Rebuilding a design culture takes years; Bloomberg underscored this by noting that Ternus had only weeks before taking the job. The indicator to watch is not hiring announcements but the consistency of the products over the next three to five years.

The risks that remain open

First, key-person risk does not disappear: it moves upward. A model that depends on the CEO's attention works as long as the CEO has it, and is fragile when the calendar fills up.

Second, artificial intelligence changes the terrain. The next relevant interface may not be a glass rectangle. One irony of history: LoveFrom, Ive's firm, now works with OpenAI, a company that aims to define the next category of devices. In addition, the delay of the Siri overhaul, which reports say will arrive with iOS 27, is a reminder that in the AI era the quality of the experience is increasingly a matter of software behavior, not just form.

Third, cost pressure. Rising memory and storage prices are pushing device prices up. A price premium holds only if perceived value keeps pace. If design stops justifying it, the advantage becomes a vulnerability.

Lessons for business leaders

Treat design as an organizational capability, not an individual's talent. What Apple lost in 2019 was not just a person but the shared standard that person embodied. Documenting principles, training successors, and rotating leadership are ways to institutionalize it.

Place design where product and pricing decisions are made. Apple's experience suggests that when design answers to an operational-efficiency function, its voice is diluted. When it answers to the CEO, it is heard.

Measure the advantage by coherence, not awards. The health signal of a design system is that each new product reinforces the others and that customers perceive them as parts of a whole.

Plan the succession of the charismatic leader before they leave. Seven years without a single leader had a cost. Companies built around a central figure should have a day-after plan starting today.

Conclusion

The question is not whether Apple can design without Jony Ive. It is whether it can do so without depending on anyone in particular. Its current answer, an engineer CEO who becomes the design chief, is consistent with its history and with its need to integrate hardware, software, and artificial intelligence more tightly. But it is also a transitional solution: it works well as long as the figure is there. The real test will be whether, a decade from now, the system still produces the same results when Ternus is no longer in the studio every day.

 

Sources consulted (press reports as of October 29, 2026; several rely on Bloomberg and Mark Gurman's Power On newsletter): Business Today (October 5, 2026), iClarified (October 4, 2026), Mactrast (January 2026), The Next Web, Outlook Business, Smartprix, and TechNews. Details about the internal governance of design come from anonymous sources cited by those outlets, not from official Apple statements