Georges Doriot

Completed for T001.5 on 2026-07-26. This synthesis is built only from the completed note files, writings.md, talks.md, and investments.md. It distinguishes Doriot's direct words, witnesses' recollections, and later interpretation.

1. Snapshot

Georges F. Doriot (1899–1987) taught Manufacturing at Harvard Business School, led wartime research in the US Army Quartermaster Corps, and served as president of American Research and Development Corporation (ARD), one of the first modern institutional venture organizations (HBS ARD history). His superpower was not a single picking formula. It was an operating system that joined technical networks, whole-company judgment, patient active ownership, and the development of people.

The record is unusually easy to mythologize. Doriot led ARD, but ARD's staff, advisers, directors, bankers, and portfolio executives originated and managed deals; individual attribution is often missing (Hsu and Kenney). ARD's reported 14.7% compound return from 1946–1971 fell to 7.4% without DEC, versus a reported 12.8% for the DJIA, so the track record is power-law evidence rather than proof of uniformly superior selection (Hsu and Kenney, p. 599).

2. Core philosophy

Doriot's most consistent distinction is between construction and speculation. Harvard reproduces his description of investing as building people and companies, while excerpts from ARD's 1971 annual report warn that capital gains had become an objective instead of a reward for useful work (HBS, “Financing New Ideas”; 1971 excerpts). This is not hostility to returns. It is a causal sequence: create a useful product, capable people, and a durable company; let value follow.

The direct written record pairs that mission with selectivity. Harvard's archive-based exhibit quotes Doriot's preference for new work that ordinary banks could not finance and for specialized companies protected by patents and know-how (HBS ARD history; HBS, “Financing New Ideas”). Yet the deals show that mission alone did not save weak economics: Island Packers reportedly lost $239,000 of a $250,000 commitment after its catch-volume assumption failed (Nicholas, p. 232).

The practical synthesis is: finance work that deserves to exist, insist that it can become an enterprise, and take responsibility for developing it.

3. Sourcing

ARD's differentiated access began with institutional networks. MIT president Karl Compton brought High Voltage Engineering; MIT expertise populated ARD's technical advisory board; DEC emerged from MIT Lincoln Laboratory (Hsu and Kenney; HBS DEC history). Doriot's HBS classroom also became a talent and referral network: Perkins, Draper, and later investors describe the course as practical operator formation (Perkins oral history; Draper oral history).

The system was not predominantly a genius investor spotting cold inbound. In the Hsu–Kenney reconstruction, internally sourced share fell from about 40% to about 20%, investment banks and brokerages supplied 25%–50% of deals in each period, unsolicited proposals rarely produced investments, and the source was unknown for up to half the deals in some windows (Hsu and Kenney, p. 594).

Actionable rule: build a network where technical truth, operating talent, and capital meet; then record attribution instead of converting institutional access into a hero story.

4. Picking

People before polished plans

The accessible record consistently favors ability, integrity, resourcefulness, and trust. Perkins remembers student teams being asked why they would trust a chosen leader with money (Perkins oral history). Aulikki Olsen says Doriot interviewed the founders' wives separately, extending diligence beyond résumés while also imposing a period-specific and intrusive family lens (Aulikki Olsen oral history).

Newhall remembers a categorical preference for scientific backgrounds, followed later by Doriot acknowledging that the rejection may have been mistaken (Newhall oral history). The useful principle is technical literacy and revisable judgment—not degree worship.

Technical defensibility with commercial practicability

Harvard reproduces Doriot's view that patents and know-how could let specialized small firms compete with large ones (HBS, “Financing New Ideas”). Nicholas describes ARD filtering for commercial practicability, patent protection, and high profit potential (Nicholas, pp. 230–231). The company, not the invention, is the unit of analysis: Doriot's 1937 course integrated products and research with capital, labor, organization, production, and administrative control (HBS, “Educating Leaders”).

Selection is strict, but not static

ARD funded roughly 2% of 6,884 evaluated proposals and never more than 4% in a five-year period (Hsu and Kenney, Table 2). Its sector and stage mix evolved: 73% of investments were R&D-intensive under the authors' proxy, while early-stage share rose from 20.6% before 1961 to 57.1% afterward (Hsu and Kenney, Table 3). This was learning under competition, not an immutable 1946 thesis.

5. Deal and portfolio mechanics

ARD mixed instruments rather than using seed equity alone: Hsu and Kenney code 34.2% equity-only and 15.0% debt-only, leaving about half in mixed or other structures (Table 3). HBS reports DEC as $70,000 of equity plus a $30,000 loan, though credible sources disagree on whether initial equity ownership was 70%, 77%, or 78% (HBS DEC history; Nicholas).

The DEC proposal shows staged ambition: begin with digital test equipment and later attempt a general-purpose computer; Nicholas says ARD added financing after milestones (1957 proposal; Nicholas, p. 234). The repeatable mechanic is to finance a bounded first proof while preserving the path to a larger technical vision.

Portfolio construction was diversified in count but concentrated in outcome. Among 99 investments with coded results, 64.6% produced a gain, only eight of the full set went public, and acquisitions were important exits (Hsu and Kenney, Tables 3–4 and n. 15). DEC's reported 1971 value dominated the economics. That is a power-law portfolio, even though Doriot framed each company as constructive work rather than a lottery ticket.

6. Working with founders

The oral histories describe active ownership that tries not to steal agency. Ken Olsen remembers ARD staying with companies through success or failure and Doriot offering ideas in ways that let the recipient own them (Ken Olsen oral history). Aulikki Olsen saw a tension: independence was expected, yet protégés were also expected to seek counsel (Aulikki Olsen oral history).

ARD's operating support could be direct. At Flexible Tubing, it reportedly installed an employee as director, treasurer, and manager; Tracerlab founder William Barbour credited Doriot with pulling him out of operational detail (Nicholas, pp. 231–232; HBS ARD history).

This was not uniformly gentle founder-friendliness. Aulikki Olsen remembers Doriot as capable of public severity, and Newhall remembers categorical screening (Aulikki Olsen; Newhall). The model is demanding mentorship, with a real risk of paternalism and overreach.

7. Views on the craft

Returns follow useful work

The 1971 annual-report excerpts object to manufacturing gains instead of products; Doriot's direct sequence is company first, capital gain second (reproduced excerpts). Yet ARD's founders also knew the model had to produce attractive economics to diffuse. The reported return record confirms why: 7.4% without DEC would not have validated the experiment against the cited 12.8% DJIA result, while 14.7% with DEC did (Hsu and Kenney, p. 599).

Vehicle design is strategy

ARD's public closed-end form gave it institutional capital but exposed it to rules that constrained affiliate transactions and employee incentives. A direct Doriot statement preserved in a congressional hearing attributes ARD's end to the Investment Company Act, though his “sole reason” should be treated as his retrospective explanation rather than complete causal proof (House hearing, pp. 339–340). Draper also remembers declining an ARD West Coast role because authority would remain in Boston (Draper oral history).

Great venture practice therefore requires alignment among capital duration, incentives, geography, and decision rights. Philosophy cannot compensate indefinitely for a vehicle that loses talent or delays action.

8. Signature bets

Digital Equipment Corporation

DEC is the fullest expression of the method: technical founders, staged product ambition, patient ownership, and a power-law result. The SEC anchors the sequence at a 1966 registration, 1968 ARD secondary, and 1970 cap-table snapshot; the remaining stock was reportedly worth $355 million when distributed in 1971 (1966 SEC digest; 1968 digest; 1970 digest; Hsu and Kenney). The widely repeated $70,000-to-$355-million comparison is roughly 5,071x gross, not an audited net multiple or an IPO value.

High Voltage Engineering

Compton sourced the MIT accelerator company; ARD reportedly invested $200,000, worth $1.8 million at the 1955 IPO (Hsu and Kenney; Gompers). It shows scientific-network access and technical distinctiveness, but the approximate 9x comparison lacks follow-on and distribution detail.

Tracerlab

Tracerlab reportedly reached sales, profit, and an IPO quickly, while its founder credited Doriot's coaching (Nicholas; HBS ARD history). It is an operating-support case, not a quantified signature return.

Flexible Tubing

ARD embedded an operator and reportedly restored profitability within a year (Nicholas, p. 232). It demonstrates intervention, though the public record lacks amount and exit.

Island Packers

The bankrupt tuna venture is a signature negative case: a technically or socially appealing project still fails when physical throughput assumptions do not work (Nicholas, p. 232).

9. Mental models and maxims

1. Constructive sequence: useful product and company first; gain as consequence (1971 excerpts).
2. Whole-company lens: judge product, people, organization, production, and capital as one system (HBS, “Educating Leaders”).
3. Able-person multiplier: prefer resourceful, trustworthy operators and develop their judgment (Perkins oral history).
4. Protected-small-company wedge: patents and know-how can neutralize incumbent scale (HBS, “Financing New Ideas”).
5. Staged technical ambition: fund a bounded proof that opens the larger product path (DEC proposal).
6. Patient active ownership: stay engaged through difficulty without reflexively selling success (Ken Olsen oral history).
7. Mentor without stealing agency: present counsel so the operator can own the decision (Ken Olsen oral history).
8. Operating arithmetic before romance: test the constraint that can kill the venture, as Island Packers illustrates (Nicholas).
9. Power-law honesty: evaluate portfolio results with and without the outlier (Hsu and Kenney).
10. Vehicle–mission fit: ensure incentives and decision rights permit patient company-building (House hearing).

Each is expanded with origin, application, and counterexample in mental-models.md.

10. Evolution and contradictions

• 1937–1966: operator curriculum. The course framed manufacturing as integrated company analysis and executive responsibility (HBS, “Educating Leaders”).
• 1946–1960: institutional experiment and cautious early mix. Only 20.6% of investments were classified early-stage, even though ARD aimed to finance new enterprise (Hsu and Kenney).
• 1957–1971: DEC validates patience and concentration. The outlier transformed ARD's reported return and the industry's imagination, but obscures weaker performance elsewhere.
• 1961–1973: more early-stage, fewer coded gains. Early-stage share rose to 57.1% while gain frequency fell, amid greater competition and incentive problems (Hsu and Kenney, Table 3).
• 1971: anti-speculation warning. Doriot argued the expanding craft was converting capital gains from consequence to purpose (1971 excerpts).
• 1978: structural retrospective. He attributed ARD's termination to the Investment Company Act (House hearing).

The deepest contradiction is between autonomy and control. Witnesses remember advice that preserved operator ownership, but also severe judgment, family diligence, categorical rules, and Boston-centered authority (Aulikki Olsen; Draper). Doriot's practice was developmental and paternal, not modern founder-friendliness.

11. Critiques and failure modes

1. Hero attribution: ARD deals are often credited to Doriot without evidence of personal sourcing or board work.
2. Survivorship and DEC dominance: the non-DEC reported return lagged the cited public benchmark (Hsu and Kenney).
3. Paternalism: family interviews, gendered assumptions, and categorical credential screens are era-bound practices, not portable maxims (Aulikki Olsen; Draper).
4. Institutional rigidity: incentive restrictions, centralized authority, and succession problems weakened the organization even as its philosophy diffused (Hsu and Kenney).
5. Mission can excuse weak economics: “noble” purpose is dangerous if it suppresses falsifiable operating assumptions; Island Packers is the warning.
6. Quote-chain risk: many famous lines are remembered decades later or reproduced through editors. The archive-only 1966 reel, 1978/1980 interviews, and speech manuscripts remain necessary checks.

12. Go deeper — ranked sources

1. Hsu and Kenney, “Organizing Venture Capital” — best portfolio and institution reconstruction.
2. Ken Olsen Smithsonian oral history — best open firsthand account of patience and mentoring.
3. DEC's May 1957 proposal — primary staged product plan.
4. HBS “Financing New Ideas” — strongest accessible Doriot excerpts on construction and technical defensibility.
5. ARD 1971 annual-report excerpts — late first-person critique; verify against the original.
6. Nicholas, “Origins of High-Tech Venture Investing” — best readable early-deal narrative.
7. Aulikki Olsen oral history — indispensable corrective on diligence, autonomy, and severity.
8. Tom Perkins oral history — Manufacturing, trust, and operator substance.
9. SEC's 1966, 1968, and 1970 DEC digests — primary chronology; follow the linked sequence in investments.md.
10. 1979 House hearing, pp. 339–340 — Doriot's structural retrospective.

Completed for T001.4 on 2026-07-26. “ARD” means American Research and Development Corporation. Doriot led the institution, but a portfolio-company investment is not automatically a personally sourced Doriot deal. Dollar outcomes below are historical reported figures, not audited fund-return disclosures.

What the record can and cannot establish

ARD evaluated 6,884 proposals and funded 120 between 1946 and 1973 in Hsu and Kenney's reconstruction of annual reports, Doriot papers, and former-officer interviews; only about 2% of evaluated proposals were funded overall, and no five-year period exceeded 4% (Hsu and Kenney, Table 2). That dataset is the best portfolio-level reconstruction found, but its authors identify missing deal-source data, differing stage-classification sources, and only 99 observations with a coded gain outcome. It is not a complete audited cash-flow ledger.

Doriot was ARD's president and the organizing influence on its practice, while its staff, technical advisers, directors, bankers, and portfolio executives originated and managed individual investments (HBS ARD history). The public evidence does not name a sourcing partner for every deal. The table therefore uses “institutional lead” for Doriot where the evidence shows ARD leadership or board involvement but not personal origination.

Deal table

This table is deliberately narrower than a portfolio list. HBS also names Cleveland Pneumatic Tool and Snyder Chemical among early controlled holdings, while the full scholarly dataset contains 120 investments; open evidence was not sufficient to report role, security, and outcome consistently for every company (CFA chapter, p. 232).

Signature wins

1. Digital Equipment Corporation — the power-law proof, with three different value dates

Olsen and Anderson's May 1957 proposal proposed selling digital test equipment before attempting a general-purpose computer. That staged product plan matters: the company did not ask investors to finance the final vision in one leap. HBS reports that ARD provided $70,000 in equity and a separate $30,000 loan, then refused acquisition approaches as the company grew (HBS DEC history).

The return story must preserve chronology:

• The SEC recorded DEC's registration as effective August 18, 1966 (SEC News Digest, August 19, 1966).
• In August 1968, ARD owned 1,750,000 of 2,296,600 outstanding shares and received permission to sell 215,000 shares; this was a secondary sale, not the IPO (SEC News Digest, August 29, 1968).
• By September 1970, after share-count changes, the SEC reported ARD at 48% of 9,675,132 shares and Ken Olsen at 10% (SEC News Digest, September 16, 1970).
• Hsu and Kenney, citing Liles and Doriot, report a $355 million gain/value when the remaining DEC stock was distributed in 1971 (Hsu and Kenney, p. 599).

Consequently, “$70,000 became $355 million” is a reported gross comparison across fourteen years. It is not the IPO value, does not include the separate loan or later financing, and does not reveal the cash timing needed for an investment-level IRR. The lesson is power-law concentration and patient support, not a clean audited 5,071x.

2. High Voltage Engineering — technical network plus an early liquidity proof

MIT president Karl Compton brought ARD the accelerator company, showing that early deal flow depended on technical institutions rather than only inbound applications (Hsu and Kenney, p. 609). Gompers reports that the $200,000 commitment was worth $1.8 million at the 1955 IPO (Gompers, p. 6). The approximate 9x gross comparison is meaningful but incomplete without dilution, follow-ons, distributions, and dates.

High Voltage also illustrates ARD's preference for technically defensible work that could be commercially unique. Hsu and Kenney classify 73% of ARD's lifetime investments as R&D-intensive and find that this classification was positively associated with a gain in their limited sample (Hsu and Kenney, Tables 3–4).

3. Tracerlab — governance mattered before a spectacular multiple existed

Tracerlab is a modest success rather than a power-law result. It reportedly reached $700,000 in sales and $30,000 in profit in 1948, then raised $1.3 million in a March IPO (CFA chapter, p. 231). HBS preserves founder William Barbour's account that Doriot helped him escape operational detail and return to long-range work (HBS ARD history). Because the board representatives identified publicly were Compton and Griswold, the evidence supports Doriot's coaching influence but not sole deal ownership.

4. Flexible Tubing — intervention as part of the investment

ARD placed an employee into three operating roles at a struggling manufacturer, which reportedly became profitable within a year after winning supply contracts (CFA chapter, p. 232). This is stronger evidence for active company-building than for financial performance: the open record provides no investment amount, exit, or multiple.

5. Ionics — a scientifically disputed bet with incomplete economics

Ionics appears in HBS's archive-based list of prized ARD companies, and Doriot served on its board (HBS ARD history). Later accounts describe outside technical skepticism of its water-purification membranes, but no primary diligence memo or verified ARD outcome was accessible. It is a signature example of technical contrarianism only with that limitation—not a quantified return case.

Misses and anti-portfolio

Realized failures

• Circo Products: contemporary TIME says ARD lost money on its first investment (TIME, 1962). The original investment instrument and final cash loss still need annual-report verification.
• Island Packers: the $250,000 co-investment ended in bankruptcy and a reported $239,000 write-off after the operating assumption about catch volume failed (CFA chapter, p. 232). This is a concrete diligence failure around a physical operating constraint.
• Portfolio loss rate: among 99 investments with coded outcomes, Hsu and Kenney report 64.6% produced a gain, implying 35.4% did not; the binary measure ignores magnitude and is incomplete for 21 of the 120 investments (Hsu and Kenney, Table 3).

Passes

TIME reported in 1962 that ARD failed to see much future in Ampex and Varian Associates, both of which later became important technology companies (TIME, 1962). The article does not disclose the proposed terms, stage, ARD decision-maker, or whether ARD formally rejected a financeable proposal. These are anti-portfolio leads, not fully reconstructed “missed returns.”

Institutional miss

ARD's later problem was not merely selecting the wrong companies. Hsu and Kenney report that the share of early-stage investments increased from 20.6% in 1946–1960 to 57.1% in 1961–1973, while the coded gain rate fell from 74.1% to 53.3% and R&D-intensive share declined (Hsu and Kenney, Table 3). They connect the deterioration to competition, sourcing, staff incentives, and ARD's public-company form rather than proving a single causal mechanism.

The clearest attribution failure would be to call every ARD investment a Doriot-sourced deal. Investment banks and brokerages supplied 25%–50% of deals in each period, internally sourced share fell, and the authors could not identify the source of up to half the deals in some five-year windows (Hsu and Kenney, p. 594).

Fund-level performance and portfolio mechanics

The power-law conclusion is supported, but it should be stated precisely: the reported portfolio return beat the cited public-market comparison only when DEC was included. Without DEC, reported compound performance lagged it. That is stronger evidence for outlier dependence than for consistent superior selection.

ARD's securities also mixed equity-only, debt-only, and combinations: Hsu and Kenney code 34.2% equity-only and 15.0% debt-only, leaving roughly half in mixed or other structures (Table 3). The modern shorthand “seed equity portfolio” erases that early risk-capital model.

Lessons grounded in the deals

1. Stage the technical vision. DEC proposed test equipment before a general-purpose computer and received additional financing after milestones (proposal; Nicholas).
2. Technical networks generated differentiated access. MIT leaders and researchers recur in High Voltage, Tracerlab, and DEC, while unsolicited proposals rarely produced investments (Hsu and Kenney).
3. Governance was part of the product. Tracerlab coaching, Flexible Tubing's embedded operator, and DEC's long holding period show post-investment work beyond financing (HBS ARD history).
4. Mission did not eliminate economics. Island Packers shows that a venture still fails on operating arithmetic; the non-DEC return shows that noble intent did not reliably generate benchmark-beating performance.
5. One outlier can validate—and conceal weaknesses in—an institution. DEC lifted the reported compound return from 7.4% to 14.7% while ARD's structure struggled to retain investment professionals (Hsu and Kenney).

Unresolved evidence requests

• Retrieve ARD's complete annual-report run and transaction ledgers to replace reported values with dated cash flows, follow-ons, dividends, write-offs, and distributions.
• Reconstruct DEC's fully diluted cap table from formation through the 1966 registration, 1968 secondary, splits, and 1971 distribution.
• Obtain investment memoranda and board minutes for Circo, High Voltage, Tracerlab, Ionics, Baird, Flexible Tubing, Ampex, and Varian.
• Separate who sourced, underwrote, sat on the board, and made the final decision for each ARD investment; do not assign team work to Doriot by default.

WRITINGS synthesis completed 2026-07-26 from the note files in notes/. The central HBS course scans are currently unavailable; this document distinguishes direct text, reproduced excerpts, and metadata-only records.

Evidence boundary

Doriot left a large written record, but most of the decisive material is in Baker Library rather than reliably accessible online. The legacy HBS links for his course notes now return 404, and their indexed Wayback snapshots could not be retrieved during this run. Notes were still created for each priority item, with exact catalog locations where available, but no arguments were inferred from titles alone. The usable online corpus is therefore a mixture of a full translated book, Harvard course-description and exhibit transcriptions, a partial edited sourcebook, a reproduced 1971 annual-report introduction, and a short 1978 interview excerpt preserved in a congressional record. HBS digital archival resources · Doriot papers finding aid

Recurring themes

1. Enterprise is constructive work, not financial manufacture

The clearest continuity runs from Doriot's teaching into his late-career criticism of venture capital. Harvard's exhibit reproduces his description of the investor's job as building people and companies, not speculating. HBS, “Financing New Ideas” In the 1971 ARD annual-report introduction, reproduced by Udayan Gupta, Doriot argued that capital gains had become an objective instead of a reward for useful work and that “manufacturing” gains was displacing manufacturing products. 1971 annual-report excerpts

This is not an anti-profit position. It is a sequencing rule: make a useful institution and product first; treat financial gain as the consequence of the constructive task. The accessible excerpts do not define acceptable holding periods, prices, or exits, so those mechanics remain for the DEALS phase. 1971 annual-report excerpts

2. Study the whole company from the operator's seat

Harvard's 1937 register describes Manufacturing as an integrated study of company analysis and administrative control. Students were to examine decisions from the perspective of the person who must make and execute them, while also understanding the viewpoints of bankers and consultants. Products, new ideas, development, and research sat beside capital, materials, labor, organization, and production control. HBS course register and exhibit transcription

The written curriculum therefore resists reducing a company to either its technology or its financing. Doriot's teaching frame was operator-centered and multi-stakeholder: understand the productive system, then judge how capital and advisers can help it. HBS, “Educating Leaders”

3. Character and judgment are operating capabilities

The accessible Harvard excerpts present leadership as a practice of observation, self-knowledge, responsibility, and constructive contribution. Doriot told students to make maximum use of both their assets and liabilities and framed an HBS education as carrying superior responsibilities as businesspeople and citizens. HBS, “Educating Leaders”

The partial preview of The First Venture Capitalist reinforces that pattern: teaching by example, developing others' judgment, and using a personal notebook as a mirror for one's thinking. It also reproduces the proposition that each investment opportunity should be considered separately. Because the sourcebook is edited and its underlying documents are not always visible in preview, these passages are strong leads rather than fully traced primary citations. Google Books preview · publisher description

The inaccessible items titled “How to Manage — Lead — Command,” “Decision Making,” the 1961 orientation class, and the 1966 closing class are precisely the documents most likely to sharpen this theme. Until their text is retrieved, the record supports a broad emphasis on character and judgment but not a detailed Doriot decision algorithm. HBS digital archival resources

4. Technical specialization can create an asymmetric opening

Harvard reproduces a 1961 Doriot speech arguing that patents and know-how can let specialized small companies compete with large organizations. HBS, “Financing New Ideas” This is the most specific accessible picking signal in the written record: seek technically protected areas where scale is not yet decisive.

The signal is narrower than a generic “invest in technology” slogan. It combines technical specialization, defensibility, and a competitive-structure claim. The original “Creative Capital” manuscript and the inaccessible 1965 “R and D” notes are needed to determine how Doriot evaluated technical novelty, market readiness, and research risk. HBS, “Financing New Ideas” · Baker Doriot papers

5. Structure can defeat philosophy

In a statement preserved in a 1979 congressional hearing record, Doriot said he terminated ARD because it could not exist under the Investment Company Act of 1940. The statement is direct but comes through an interview, a legal memorandum, and then the official hearing record. House hearing record, printed p. 339 n. 12

The writing supplies an institutional-design lesson: patient company-building requires a vehicle whose regulation and compensation rules can retain investors and operators. Doriot's “sole reason” wording records his stated rationale; it does not by itself prove that ARD's merger had no other causes. House hearing record, printed pp. 339–340

Evolution of the written record

• 1928 — intellectual context, not doctrine. Doriot co-translated Henri Sée's Modern Capitalism. The full text treats capitalism as a historically evolving combination of commercial, financial, and industrial systems, but Sée authored the argument. Doriot's translation role is evidence of exposure, not proof that he adopted every proposition. Google Books
• 1937–1966 — the company-building curriculum. The course register and later Harvard excerpts show a shift toward integrated company analysis, executive judgment, innovation, character, and civic responsibility. Most of the exact course texts remain retrieval targets. HBS, “Educating Leaders” · HBS digital archival resources
• 1961–1967 — creative capital. Accessible excerpts connect technical defensibility with small-company opportunity and contrast construction with speculation. HBS, “Financing New Ideas”
• 1971 — warning against speculative drift. Doriot's annual-report introduction says the expanding industry was forgetting the constructive task and promoting gains from outcome to goal. Reproduced annual-report introduction
• 1978 — structural retrospective. His surviving interview excerpt attributes ARD's loss of independence to the 1940 Act, moving the written emphasis from how to build companies to whether the investment vehicle permits that work. House hearing record

Five best pieces to read first

1. ARD 1971 annual-report introduction — the clearest accessible first-person contrast between constructive venture work and speculation, though only reproduced excerpts are online. Read the reproduction
2. The First Venture Capitalist — the broadest edited collection of Doriot-attributed material on leadership, capital, research, and organization; use the full edition and trace each passage to its origin. Publisher
3. 1937 HBS course register — a compact map of the operator-centered curriculum from which Doriot's later investment method grew. Harvard exhibit transcription
4. 1978 Doriot interview excerpt — narrow but direct evidence that investment-vehicle structure was central to his explanation of ARD's end. House hearing record
5. 1966 closing statement to the Manufacturing class — probably the highest-value unread primary item because it closes four decades of teaching; the online scan is unavailable, so request Baker volume 126, item 35. Baker finding aid

Open questions and retrieval queue

1. Retrieve the 1966 closing statement, 1961 orientation, “How to Manage — Lead — Command,” “Decision Making,” “What Is a Company?”, and “R and D” from Baker Library; do not rely on their dead legacy links.
2. Obtain the original 1971 ARD annual report to verify the reproduced wording, punctuation, authorship, and omitted context.
3. Locate Paul Dykstra's full 1978 interview transcript and test whether Doriot qualified the “sole reason” explanation.
4. Trace each useful passage in The First Venture Capitalist to its original document and date.
5. Determine whether Doriot ever cited Sée's framework after translating Modern Capitalism; absent that link, keep it as intellectual context only.

Completed for T001.3 on 2026-07-26. This synthesis separates authenticated Doriot recordings and manuscripts from eyewitness recollections and modern retellings. A catalog record proves that an item exists; it does not prove what the inaccessible item says.

Evidence and access boundary

The public record is rich in testimony about Doriot but poor in publicly playable Doriot audio. Harvard catalogs a March 1, 1966 classroom reel, a 17-page 1980 interview transcript, and a 21-minute 1988 documentary, while the Computer History Museum catalogs a 1978 interview transcript and birthday/documentary production reels. None was publicly readable or streamable during this run. They are retrieval targets, not evidence for a paraphrase or quotation.

Four cataloged speech manuscripts are similarly archive-only: “Investing in New Areas of Science”, “Venture Capital”, “Launching New Products vs. Launching a New Company”, and “The Role of the Entrepreneur”. Their titles help prioritize retrieval but cannot bear claims about Doriot's frameworks.

The strongest open evidence is therefore firsthand recollection: approved oral histories with Aulikki Olsen, Ken Olsen, Tom Perkins, William Draper III, and Charles Newhall. These witnesses knew Doriot, but they spoke years later and sometimes quote him from memory. Modern podcasts by Spencer Ante, Margaret O'Mara, and David Senra are useful syntheses, not substitutes for ARD filings, board records, or Doriot's own papers.

Recurring stories

Venture capital is mentorship plus money

Margaret O'Mara describes high-technology venture capital as money combined with mentorship and expertise in the timestamped “Venture Capital Built Silicon Valley” transcript. The firsthand record supports that distinction. Ken Olsen remembers Doriot as “always there as a mentor and for help” in the Smithsonian oral history. Perkins calls Doriot his second mentor and recalls Manufacturing as practical formation in making, selling, and satisfying customers in his Bancroft Library oral history. Draper remembers advice to keep alert to opportunity and then acquire operating experience in his Computer History Museum oral history.

The recurring model is not passive friendliness. Aulikki Olsen remembers whole-family diligence, explicit boundaries between family and company, and a mentor who could be both kind and severe in her approved oral-history transcript. Newhall remembers being rejected for lacking scientific training, then receiving a later acknowledgment that the judgment may have been wrong in his CHM oral history. Mentorship included screening, challenge, correction, and revision.

Build enduring companies, not quick trades

Ken Olsen supplies the clearest eyewitness account of the time horizon. He says ARD stayed with companies through success or failure and remembers Doriot warning that unusually early success could endanger survival in the Smithsonian transcript. Spencer Ante summarizes the same orientation as patient company-building rather than speculation at 22:34–25:38 of the Libertarianism.org transcript. The official notes for Founders #203 repeat the line, “I am building men and companies,” but that quotation remains a secondary chain until traced to Doriot's original document.

This theme is stronger than a slogan because witnesses describe operating mechanisms: patient ownership, advice presented without stealing the operator's agency, experiential student projects, and attention to the company's ability to survive. It is also incomplete without the harder edge recorded by Olsen and Newhall: Doriot could be categorical about people, background, and board rules.

Teaching, wartime R&D, and investing formed one operating system

Ante links Doriot's wartime responsibility for research, manufacturing, and product development to his later investment practice in the 2020 podcast transcript. Perkins and Draper describe the peacetime teaching version: teams selected leaders, divided work, studied real companies or industries, generated ideas repeatedly, and were asked why they trusted particular people (Perkins; Draper).

The combined evidence suggests a consistent practice: learn through consequential work, judge the people doing it, and make the organization stronger through feedback. That is an inference across witnesses, not a recovered Doriot checklist. The inaccessible 1966 class recording is the best test of whether the remembered method matches Doriot's actual classroom language.

Institutional design determines behavior

The open talks also explain why Doriot's philosophy cannot be reduced to founder coaching. Ante describes ARD's disputes with the SEC over valuations and stock options, then connects later limited-partnership structures to stronger talent incentives in the 2020 transcript. Draper recalls declining an ARD West Coast role because Boston would retain decisions while he only sourced opportunities in his oral history. His account exposes a centralization problem inside the pioneering vehicle.

The recurring lesson is two-sided: Doriot helped institutionalize patient, expert venture investing, yet ARD's public-company structure and centralized authority constrained incentives and geographic expansion. These retrospective judgments need testing against ARD records in the DEALS phase.

Repeated frameworks

1. Construct, do not speculate. The consistent distinction is productive company-building versus trading in financial claims. It appears in Ante's summary, the Founders show notes, and Perkins's recollection of making, selling, and satisfying customers.
2. Back and develop people. Technical literacy, integrity, judgment, resourcefulness, and the ability to earn trust recur more often than a fixed market-size formula (Newhall; Perkins).
3. Use patient, active ownership. Ken Olsen remembers support continuing through success or failure rather than a reflexive sale (Smithsonian).
4. Preserve operator agency while remaining available. Ken and Aulikki Olsen both describe independence combined with access to advice, though Aulikki explicitly limits her knowledge of the business relationship (Aulikki Olsen).
5. Create strength below the leader. The Founders notes emphasize intergenerational teams and capability beneath the founder; the student-team recollections show an early teaching analogue (Founders #203).
6. Match authority to responsibility. Draper's account of a West Coast office without decision rights shows the failure mode when this principle was not met.

Claims said once, contested, or easy to overstate

• The DEC value chronology is often misstated. Ante says the $70,000 investment exceeded $400 million “when Digital went public”; Founders repeats a more-than-$400-million outcome without a measurement date. SEC records establish that DEC's registration became effective on August 18, 1966, while the much larger value normally refers to a later ARD distribution. The podcasts should not be used for IPO value or realized-return timing.
• Initial ownership and financing bases vary. Oral histories and later summaries use $70,000, while broader accounts sometimes add a separate loan or report stakes around 70%, 77%, 78%, or 80%. Each figure needs a dated cap table and transaction definition.
• “First venture capitalist” is definitional. Perkins carefully says first US high-technology venture capitalist “of the new mold”; Stanford and Ante emphasize institutional form. Earlier risk capital existed, so the defensible claim concerns a particular postwar institutional model, not the first risky investment.
• Memorable quotations are often remembered, not recorded. The Olsen, Perkins, Draper, and Newhall transcripts preserve valuable firsthand testimony, but their Doriot quotations are recollections decades after the events. The Founders quotation adds another layer through a biography and podcast.
• Mentorship had a severe side. Aulikki Olsen's account of favored and disfavored people and Newhall's remembered technical-degree screen complicate a purely nurturing canon. These are important counter-signals, not grounds for universalizing one anecdote.

Ranked listen/read-first list

1. Ken Olsen oral history — best open firsthand account of patience, survival, mentoring style, and the DEC relationship.
2. Aulikki Olsen oral history — strongest corrective on whole-person diligence, family boundaries, autonomy, loyalty, and severity.
3. Tom Perkins oral history — clearest description of Manufacturing as operator formation and trust-based leadership training.
4. Venture Capital Built Silicon Valley — best open, timestamped secondary synthesis; use its institutional history but reject its compressed DEC value chronology.
5. 1966 class recording — highest-value archival retrieval because it can replace remembered classroom language with authenticated Doriot audio.

Retrieval queue

Request the 1966 class reel and the 1980 Doriot interview from Baker Library first, then the 1978 interview and production reels from CHM. Retrieve the four speech manuscripts as one archival batch. For each item, preserve exact page or timestamp anchors, identify whether Doriot approved the transcript, and distinguish prepared remarks from audience questions or editorial narration.

Completed for T001.5 on 2026-07-26. Direct quotations are labeled separately from remembered Doriot speech and later summaries. A remembered line is evidence about a witness's experience, not a stenographic transcript.

1. Constructive sequence

Statement: Build a useful product, capable people, and a durable company before treating financial gain as the objective.

Origin: Doriot's 1971 ARD introduction warned that the industry was “manufacturing capital gains” rather than manufacturing products (reproduced annual-report excerpts).

When it applies: Use it to test incentives, board discussions, and exit pressure: is the proposed action strengthening the productive enterprise or merely improving a near-term mark?

Counterexample / limit: A useful company still needs economic discipline. Island Packers' reported $239,000 write-off shows that mission does not repair a fatal operating assumption (Nicholas, p. 232).

2. Whole-company lens

Statement: Evaluate technology, product, organization, production, people, capital, and control as one operating system.

Origin: The 1937 Manufacturing course description joined products and research with “capital, materials, labor, organization and production control” (HBS exhibit transcription).

When it applies: Use it when a technically impressive pitch treats commercialization, hiring, or production as somebody else's problem.

Counterexample / limit: An integrated lens can become diffuse. Assign explicit owners and falsifiable milestones rather than conducting endless holistic analysis.

3. Able-person multiplier

Statement: An operator's judgment, integrity, resourcefulness, and ability to earn trust can multiply an imperfect idea.

Origin: Perkins remembers Doriot asking why the students chose the person they “would trust your money with” (Perkins oral history, transcript p. 6).

When it applies: Use it in founder references, leadership succession, and moments when the plan must change but the team must keep learning.

Counterexample / limit: Person-first investing can become charisma bias. Newhall's remembered rejection of an English major shows how a human heuristic can harden into a crude credential screen (Newhall oral history, p. 19).

4. Protected-small-company wedge

Statement: A specialized small company can neutralize incumbent scale when patents and know-how protect a valuable technical wedge.

Origin: Harvard reproduces Doriot's argument that “patents and know-how” could let small specialized companies compete with large organizations (HBS, “Financing New Ideas”).

When it applies: Use it where technical knowledge is difficult to reproduce, the initial market values performance, and a small team can reach customers before incumbents mobilize.

Counterexample / limit: Protection without a commercially practicable product is not a business. ARD's own filters reportedly included commercial practicability and high profit potential, not patents alone (Nicholas, pp. 230–231).

5. Staged technical ambition

Statement: Finance a bounded first proof that creates information and preserves the route to a much larger vision.

Origin: DEC's proposal began with digital test equipment and positioned a general-purpose computer as the next phase (May 1957 proposal).

When it applies: Use it when the final product requires too much capital or technical risk to fund in one step; define a first product that has its own customers and de-risks the platform.

Counterexample / limit: A small first product can become a local optimum. The milestone must test the core technical or commercial thesis, not merely generate unrelated revenue.

6. Patient active ownership

Statement: Stay with a company through difficulty and success while actively helping it become stronger.

Origin: Ken Olsen remembers that ARD would “work with, the company until they were successful, or until they failed” (Smithsonian oral history).

When it applies: Use it when the underlying thesis remains valid but development, hiring, or commercialization takes longer than expected.

Counterexample / limit: Patience without an explicit kill test becomes denial. Doriot's “family” framing must not prevent recognizing when the thesis is falsified.

7. Mentor without stealing agency

Statement: Offer experienced counsel in a form the operator can understand, adapt, and own.

Origin: Ken Olsen says Doriot “was always there as a mentor and for help” and usually presented ideas without demanding credit (Smithsonian oral history).

When it applies: Use it in board work, founder coaching, and succession development where durable judgment matters more than winning one decision.

Counterexample / limit: Aulikki Olsen remembers the tension between independence and being expected to seek Doriot's advice; mentorship can become dependency or paternal control (oral history, p. 9).

8. Operating-constraint kill test

Statement: Identify the physical, technical, regulatory, or unit-economic constraint that can make every other strength irrelevant.

Origin: Island Packers reportedly went bankrupt because it could not catch enough fish to sustain the business, producing a $239,000 write-off on ARD's $250,000 commitment (Nicholas, p. 232).

When it applies: Use it before investment and at each follow-on: name the one assumption whose failure would destroy the model, then test it directly.

Counterexample / limit: Not every uncertainty is a kill constraint. Overweighting one measurable risk can blind the investor to adaptive founders or alternative markets.

9. Power-law honesty

Statement: Judge a venture portfolio both with and without its largest outlier.

Origin: Liles's reported ARD return is 14.7% with DEC and 7.4% without it, versus a cited 12.8% DJIA compound return (Hsu and Kenney, p. 599).

When it applies: Use it in fund attribution, reserve planning, and performance review to distinguish the success of the portfolio construction from repeatable company selection.

Counterexample / limit: Removing the outlier is diagnostic, not a claim that it does not count. Venture portfolios are built to own rare extreme outcomes; the honest analysis reports both.

10. Vehicle–mission fit

Statement: Capital duration, incentives, governance, and decision rights must support the investment behavior the strategy requires.

Origin: Doriot later said ARD “could not exist under the 1940 Act,” a direct statement preserved through an interview and congressional hearing record (House hearing, printed pp. 339–340).

When it applies: Use it when choosing fund structure, compensation, geography, and delegation: can the vehicle retain talent, act quickly, and remain patient?

Counterexample / limit: Regulation was not necessarily the sole cause of ARD's decline. Competition, centralized authority, succession, and Doriot's own management choices also appear in the scholarly account (Hsu and Kenney).

Compact decision checklist

Before investing:

1. What useful enterprise is being constructed?
2. Why is this operator trustworthy, adaptive, and technically credible?
3. What protects the small company's wedge?
4. What bounded first proof unlocks the larger vision?
5. Which operating constraint can kill the thesis?

After investing:

1. Is the company becoming stronger, or only better marked?
2. Does counsel increase founder judgment or replace it?
3. Is patience still thesis-based?
4. Does this result survive portfolio analysis without the outlier?
5. Does the vehicle permit the behavior the mission requires?

T001.6 completed 2026-07-26. Scope: citation support, link resolution, deal attribution, return chronology, quote integrity, access honesty, and cross-investor handoff.

Result

Pass after corrections. The core thesis survives: Doriot's repeatable contribution was a company-building system, while DEC supplied the economic outlier. Three unsupported board-role details were removed, one partner-attribution citation was made local to the claim, two unsupported SCOUT-stage company leads were removed, and the index description was narrowed.

Citation reopenings

Five deal-attribution cross-checks

Mechanical and evidence QA

• sources.md has 105 annotated entries, 50 marked processed, and zero duplicate URLs.
• All 36 w-* and t-* note files use the required template and remain below 500 words.
• No quoted passage in the note corpus exceeds 25 words.
• investmentphilosophy.md contains all 12 required sections; mental-models.md contains 10 models with origin, use, and counterexample.
• Archive-only audio, interviews, manuscripts, and film remain labeled partial; no listening or reading is implied.
• DEC's 1966 registration, 1968 secondary, 1970 ownership snapshot, and reported 1971 distribution remain separate everywhere reviewed.

Residual gaps

The work is shareable but not archival-final. A full ARD annual-report and cash-flow reconstruction could revise company amounts and return timing. The 1966 class recording, 1978 and 1980 Doriot interviews, four speech manuscripts, and original 1971 annual report remain the highest-value retrievals.

SCOUT inventory accessed 2026-07-26. Priority: ★★★ must process · ★★ should process · ★ if time. Reliability: A original document, filing, or direct words · B strong institutional, scholarly, journalistic, or firsthand-secondary source · C lead only.

Primary writings and course documents

• ★★★ | Closing Statements to the Manufacturing Class | 1966-05-17 | speech/course document | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; Baker volume 126, item 35 is the next legitimate retrieval route.
• ★★★ | Orientation Class for First-Year Students | 1961-05-05 | speech/course document | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata plus a short Harvard exhibit excerpt; request the Baker original.
• ★★★ | How to Manage — Lead — Command | 1959-04-13 | course notes | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; Baker volume 5, folders 47–48 require date reconciliation and retrieval.
• ★★★ | Decision Making | 1959-04-14 | course notes | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; no framework was inferred without the text.
• ★★ | Notes on the Duties of the Assistant to Professor Doriot | 1952-06 | memo | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; Baker box 1, folder 6.1 is the next retrieval route.
• ★★ | George E. McCown Assistant's Lecture | 1962-10 | course document | reliability: B | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; Baker box 1, folder 11.1 is the next retrieval route.
• ★ | Manufacturing I Final Exam | 1964-12-18 | exam | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata only; no questions or grading criteria were inferred.
• ★ | Manufacturing II Final Exam | 1965-03-06 | exam | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata only; no questions or grading criteria were inferred.
• ★ | Special Speakers for Doriot's Class | 1937–1938 | course record | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Maps the practitioner network feeding the Manufacturing course.
• ★ | Talk by General Leslie Groves | 1958-12-08 | course notes | reliability: B | access: partial/fragile, archived snapshot indexed but not retrieved | Context for Doriot's use of high-stakes technical leadership cases.
• ★★★ | R and D | 1965 | course notes | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; Baker volume 4, folder 15 is the next retrieval route.
• ★★ | Advertising — Publicity — Public Relations | 1958; rev. 1964 | course notes | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; Baker volume 5, folder 26 is the next retrieval route.
• ★★★ | What Is a Company? | undated | course notes | reliability: A | access: partial/fragile, archived snapshot indexed but not retrieved | Processed as metadata; related secondary material remains an unverified lead.
• ★★★ | Digital archival resources index | 1927–1968 collection | digital collection | reliability: B | access: full index, mixed document access | Processed as the canonical metadata map for dead course-document links.
• ★★ | Official Register of Harvard University: Graduate School of Business Administration | 1937 | institutional register | reliability: A | access: partial, Harvard exhibit transcription of p. 65 | Processed for the course description; the scan viewer did not render.
• ★★ | Georges F. Doriot: Manufacturing Class Notes | 1927–1966; published 1993 | edited primary compilation | reliability: A | access: partial, catalog/library access | Processed as metadata only; obtain the compilation before quoting.
• ★★ | The First Venture Capitalist | 2004 | edited book | reliability: B | access: partial, preview and publisher description | Processed from visible excerpts; trace each passage to the original papers.
• ★ | Modern Capitalism: Its Origin and Evolution | 1928 | translated book | reliability: A | access: full public-domain reprint | Processed as intellectual context; Sée's arguments were not attributed to Doriot.

Archives and finding aids

• ★★★ | Georges F. Doriot papers, 1920–2002 | 1920–2002 | archival collection, 39 linear feet | reliability: A | access: partial, reading room/finding aid | Teaching notes, correspondence, writings, speeches, ARD records, board material, annual reports, and a 1980 interview transcript.
• ★★★ | American Research and Development papers, 1946–1991 | 1946–1991 | archival collection | reliability: A | access: partial, reading room/finding aid | ARD annual reports, Doriot address and advice files, CED/EED records, and the 1988 film.
• ★★★ | Georges F. Doriot papers — Library of Congress | 1917–1985 | archival collection | reliability: A | access: partial, reading room/finding aid | 3,600 items: diaries, correspondence, reports, lectures, photographs, and wartime R&D files.
• ★★ | Library of Congress finding-aid PDF | updated 2023 | finding aid | reliability: A | access: full | Container-level chronology and route to specific military, lecture, and diary materials.
• ★★ | Georges F. Doriot Research Collection, 1928–1952 | 1928–1952 | archival collection | reliability: A | access: partial, reading room/finding aid | Research on aviation, atomic energy, radioisotopes, and correspondence behind the Manufacturing course.
• ★★ | Kenneth H. Olsen collection, 1940–2014 | 1940–2014 | archival collection | reliability: A | access: partial/restricted, physical and born-digital records by request | Doriot–Olsen relationship, DEC–ARD governance, and Olsen's service on ARD's board.
• ★★ | Patrick R. Liles papers, 1935–1981 | 1935–1981 | archival collection | reliability: A | access: partial, finding aid | Interviews and original company files used to analyze ARD and early venture-firm durability.
• ★★ | Digital Equipment Corporation records | 1947–1998 | archival collection | reliability: A | access: partial, finding aid | Contains Doriot-related material and DEC governance/company records.
• ★★ | William Elfers collection of Georges F. Doriot letters | 1956–1984 | letters and photographs | reliability: A | access: partial, reading room | Five direct later-life letters on ARD, Greylock, business, and Doriot's outlook.
• ★★ | Kenneth and Aulikki Olsen interviews about Doriot | 2003 | four audio files plus ASR transcripts | reliability: B | access: partial, by request | Extended eyewitness accounts of Doriot's mentoring, character, and relationship with DEC; ASR errors require audio verification.
• ★ | HBS archival-collections map | collection page | archival guide | reliability: B | access: full | Crosswalk among Baker, Library of Congress, DEC, and related venture-capital collections.

Video, audio, oral history, and transcript-bearing sources

• ★★★ | Class Recordings — Prof. Georges F. Doriot | 1966-03-01 | open-reel classroom audio | reliability: A | access: partial, Harvard request | Processed as metadata only; no content was inferred without the reel or a transcript.
• ★★★ | Georges F. Doriot—Win Hindle papers | 1971–1987 | archival file with 1978 interview transcript | reliability: A | access: partial, CHM request | Processed as metadata only; request the transcript and inspect approval/provenance.
• ★★★ | Aulikki Olsen interview with General Georges F. Doriot | 1980 | 17-page typed transcript | reliability: A | access: partial, Harvard request | Processed as metadata only; Baker retrieval is required before summarizing Doriot's answers.
• ★★★ | Doriot 80th-birthday and documentary production reels | 1979–circa 1987 | film, audio, and VHS archive | reliability: A/B | access: partial, CHM request | Processed as a segment-level retrieval target; speakers and editorial layers remain unidentified.
• ★★★ | Investing in New Areas of Science | 1961-03-09 | speech manuscript | reliability: A | access: partial, Harvard request | Processed as metadata only; no “creative capital” content was assumed from the title.
• ★★★ | Venture Capital — HBS Club of Toronto | 1966-09-06 | speech manuscript | reliability: A | access: partial, Harvard request | Processed as metadata only; retrieve Volume 126, Item 36.
• ★★ | Launching New Products vs. Launching a New Company | 1968-02-08 | speech manuscript | reliability: A | access: partial, Harvard request | Processed as metadata only; the title cannot establish the comparison.
• ★★ | The Role of the Entrepreneur | 1973-06-07 | speech manuscript | reliability: A | access: partial, Harvard request | Processed as metadata only; founder-selection claims await the manuscript.
• ★★★ | Georges F. Doriot: Man With a Vision | 1988 | 21-minute documentary | reliability: B | access: partial, Harvard request | Processed as metadata and a source-discovery target; film and credits remain inaccessible.
• ★★★ | Aulikki Olsen oral history | 1991-03-05 and 1991-06-18 | approved oral-history transcript | reliability: B | access: full | Processed for whole-person diligence, boundaries, autonomy, loyalty, and explicit witness limitations.
• ★★★ | Ken Olsen oral-history transcript | 1988-09-28–29 | Smithsonian oral history | reliability: B | access: full | Processed for patient ownership, survival discipline, mentoring, and DEC-proposal recollections.
• ★★ | Tom Perkins oral history | 2009 | oral-history transcript | reliability: B | access: full | Processed for Manufacturing pedagogy, operator substance, leadership trust, and retrospective ARD critique.
• ★★ | William H. Draper III oral history | 2011-04-14 | oral-history transcript | reliability: B | access: full | Processed for alertness, team learning, operating experience, and ARD's centralization tension.
• ★★ | Charles W. Newhall III oral history | 2013-04-18 | oral-history transcript | reliability: B | access: full | Processed for technical screening, revisable judgment, and the remembered NEA naming story.
• ★★★ | A History of Venture Capital | 2009-02-04 | Stanford talk/podcast | reliability: B | access: partial review, full audio available | Processed from the official synopsis and 2:27 excerpt; the claimed full transcript is only one page.
• ★★★ | Venture Capital Built Silicon Valley | 2020-02-06 | podcast and authored timestamped transcript | reliability: B | access: full | Processed for mentorship, wartime formation, company-building, SEC constraints, and a corrected DEC-value chronology.
• ★★ | Founders #203 — Georges Doriot | 2021-09-08 | podcast | reliability: B | access: partial review, full audio available and official transcript login-gated | Processed from official show notes and preview; quotations and return claims require primary reconciliation.

Books and long-form research

• ★★★ | Creative Capital: Georges Doriot and the Birth of Venture Capital | 2008 | biography | reliability: B | access: partial, table of contents and chapter-eight opening | Processed only for accessible preview content; obtain full lawful access before broader use.
• ★★★ | VC: An American History | 2019 | scholarly history | reliability: B | access: partial preview | Places ARD in the longer history of risk capital and challenges simple “first VC” claims.
• ★★★ | The Origins of High-Tech Venture Investing in America | 2016 | scholarly chapter | reliability: B | access: full | Processed for early-company amounts, operating outcomes, DEC staging, portfolio structure, misses, and performance caveats.
• ★★ | Done Deals | 2000 | oral-history anthology | reliability: B | access: partial | Charles Waite and other practitioners describe ARD and the formation of the venture industry.
• ★★ | The Ultimate Entrepreneur | 1988 | company biography | reliability: B | access: full | Detailed Ken Olsen/DEC account with Doriot governance and founder-support evidence; verify rights before quoting.
• ★ | A Delicate Experiment | 1987 | institutional history | reliability: B | access: partial | HBS context for Doriot's course and the school's pre-1945 development.
• ★ | Sustaining the Venture Capital Firm | 1977 | research monograph | reliability: B | access: partial | Early practitioner research using interviews and ARD files; source for performance and organizational claims.
• ★★ | Processing the Future: Venture Project Evaluation at ARD | 2018 | scholarly chapter | reliability: B | access: partial, abstract | Focused analysis of how ARD evaluated uncertain technical ventures and converted narratives and forecasts into decisions.
• ★ | Georges F. Doriot: Dream Builder | 2015 | book chapter | reliability: B | access: partial | Concise later synthesis with explicit attention to ARD formation and major deals.

Biographical and institutional context

• ★★★ | HBS Baker Library — introduction | 2014 exhibit | institutional biography | reliability: B | access: full | Authoritative overview and gateway to Doriot's four major careers.
• ★★ | Early Years at HBS | 2014 exhibit | institutional biography | reliability: B | access: full | Education, early teaching, and intellectual formation.
• ★★★ | Educating Leaders | 2014 exhibit | institutional biography | reliability: B | access: full | Processed for archival excerpts on company-building, character, and responsibility.
• ★★★ | Wartime Innovation | 2014 exhibit | institutional biography | reliability: B | access: full | How Army R&D shaped his later venture method.
• ★★★ | ARD | 2014 exhibit | institutional biography | reliability: B | access: full | Processed for team roles, named early holdings, board participation, Tracerlab coaching, and archive-based source leads.
• ★★★ | DEC | 2014 exhibit | institutional biography | reliability: B | access: full | Processed for proposal history, the $70,000 equity plus $30,000 loan account, long holding period, and ownership-figure conflict.
• ★★★ | Financing New Ideas | 2014 exhibit | institutional biography | reliability: B | access: full | Processed for Doriot excerpts on technical defensibility, construction, and speculation.
• ★★ | CPA and INSEAD | 2014 exhibit | institutional biography | reliability: B | access: full | Doriot's European management-education project and cross-border network.
• ★ | The French Library | 2014 exhibit | institutional biography | reliability: B | access: full | Cultural institution building and the role of Edna Doriot.
• ★★ | Visionary, Innovator, Educator | 2014 exhibit | institutional biography | reliability: B | access: full | Legacy assessment grounded in Baker holdings.
• ★★★ | HBS Entrepreneurship profile | current profile | institutional biography | reliability: B | access: full | HBS teaching chronology and concise ARD/DEC narrative.
• ★★ | HBS 20th-Century Leaders profile | database profile | institutional biography | reliability: B | access: full | Useful cross-check, but its 80% DEC figure conflicts with other sources.
• ★★ | Baker Library exhibit announcement | 2014 | institutional news | reliability: B | access: full | Quotes Doriot's ARD language and summarizes teaching, Army, and INSEAD roles.
• ★★ | Baker Library receives the Doriot papers | 2012-04-23 | institutional news | reliability: B | access: full | Describes collection scope, dates, speeches, annual reports, and personal library.
• ★★★ | INSEAD — A New Vision, 1950–1959 | institutional history | reliability: B | access: full | Official founding timeline and Doriot's postwar rationale.
• ★★ | INSEAD at 50: The Defining Years | 2009-10-22 | institutional history | reliability: B | access: full | Gives the school's account of Doriot's two postwar projects; its DEC return needs correction/triangulation.
• ★★ | French Library history | institutional history | reliability: B | access: full | Doriot's 1978–1987 presidency and the Doriots' civic work.
• ★★★ | Status Reports of Quartermaster R&D Projects | 1942-09–1943-01 | five official monthly reports | reliability: A | access: partial/fragile, indexed but unavailable during QA | Direct documentary record of the portfolio, cadence, and practical operating system of Doriot's wartime R&D branch.
• ★★ | Wound Ballistics — Personnel Protective Armor | 1962 | official Army medical history | reliability: B | access: full | Official reconstruction of the Doriot-led “Doron” body-armor program and its experimental lineage.
• ★★★ | U.S. Army official history — Quartermaster Corps | official history | reliability: B | access: full | Institutional context for Doriot's wartime R&D organization.
• ★★ | Quartermaster Hall of Fame — BG Georges F. Doriot | institutional biography | reliability: B | access: full | Focused military biography and postwar “Institute of Man” work.
• ★★ | Smithsonian National Portrait Gallery record | collection record | reliability: B | access: full | Exact life dates and identity metadata.
• ★★ | Who Made America? — Georges Doriot | PBS profile | reliability: B | access: full | Accessible biography linked to a researched documentary series; useful lead, not final authority for returns.

Deals, fund structure, performance, and criticism

• ★★★ | ARD notice and order for hearing, File 812-440 | 1946-07-16 | Federal Register filing | reliability: A | access: full | Foundational regulatory record for ARD's institutional ownership, non-diversified status, and public offering.
• ★★★ | Organizing Venture Capital: The Rise and Demise of ARD | 2005 | peer-reviewed paper | reliability: B | access: full | Processed for the 120-investment reconstruction, sourcing, stage/sector shifts, gain data, fund-return estimates, and organizational failure.
• ★★★ | The Rise and Fall of Venture Capital | 1994 | scholarly article | reliability: B | access: full | Processed for High Voltage's reported 1955 value, DEC's power-law role, and the hands-on ARD model.
• ★★★ | A Proposal to American Research and Development Corporation | 1957-05-27 | company proposal | reliability: A | access: full | Processed for DEC's staged test-equipment-to-computer product plan; the scan's financial OCR remains difficult.
• ★★★ | Digital Equipment registration effective, File 2-25291 | 1966-08-19 | SEC News Digest | reliability: A | access: full | Processed to establish the 18 August 1966 registration date.
• ★★★ | ARD permitted to sell 215,000 DEC shares | 1968-08-29 | SEC News Digest | reliability: A | access: full | Processed to separate the 1968 secondary from the IPO and anchor ARD's then-current holding.
• ★★ | DEC 500,000-share offering, File 2-38367 | 1970-09-16 | SEC News Digest | reliability: A | access: full | Processed as a later cap-table snapshot before the reported 1971 distribution.
• ★★ | ARD application to cease being an investment company | 1973-03-05 | SEC News Digest | reliability: A | access: full | Primary closeout evidence after the Textron transaction.
• ★★ | ARD qualifying certification | 1966-02-11 | SEC News Digest | reliability: A | access: full | Defines ARD's closed-end, non-diversified structure, purpose, and tax/regulatory status.
• ★★ | ARD and Tracerlab affiliate-transaction application | 1957-01-11 | Federal Register filing | reliability: A | access: full | Processed as a later Tracerlab ownership and affiliate-transaction snapshot, not initial-deal terms.
• ★★ | ARD and Baird Associates application | 1953-04-18 | Federal Register scan | reliability: A | access: full | Processed as primary later-stage Baird ownership and recapitalization evidence, not a return source.
• ★★★ | Small Business Investment Incentive Act hearings, Part 2 | 1979-11-07 | congressional hearing | reliability: A | access: full | Processed for the Doriot interview excerpt at printed p. 339 n. 12.
• ★★★ | General Doriot and Ken Olsen | 1957 event / CHM artifact | reliability: B | access: full | CHM's DEC photo record reports $70,000 for 70% and preserves a Doriot statement on “thought-out risk.”
• ★★★ | Digital This Week — In Memoriam: General Georges F. Doriot | 1987-06-15 | company newsletter | reliability: B | access: full | Contemporaneous DEC account of his board service and patient-investing influence.
• ★★ | DEC first board of directors | circa 1960 | company photograph/archive | reliability: A | access: full | Identifies ARD representatives on DEC's first board; guards against wrongly crediting Doriot personally.
• ★★ | Digital Equipment Corporation: The First Twenty-five Years | 1982 address, published 1983 | founder address/company history | reliability: A | access: full | Ken Olsen's account of the proposal, ARD's objections and advice, the $70,000 equity financing, early profit, and Doriot's role.
• ★★ | HBS Ken Olsen collection announcement | 2019-11-15 | institutional news | reliability: B | access: full | Confirms Olsen–Doriot relationship, ARD board materials, and Dorothy Rowe's DEC role.
• ★★ | The Consequences of Entrepreneurial Finance | 2010 | scholarly working paper | reliability: B | access: full | Comparative empirical context for ARD's organizational form and entrepreneurial finance.

Contemporary and retrospective press

• ★★ | Industry: Profit-Minded Professor | 1962-03-02 | contemporary magazine profile | reliability: B | access: full article text via publisher index | Added and processed for Circo's first-investment loss, Ampex and Varian pass leads, and contemporary limits on ARD's record.
• ★★★ | General Doriot's Dream Factory | Fortune, 1967-08 | magazine profile | reliability: B | access: partial, catalog metadata | Major contemporary profile containing Doriot's construction-versus-speculation distinction and outside criticism; obtain article pages before quoting.
• ★★★ | HBS Doriot bibliography | sources dated 1927–2012 | institutional bibliography | reliability: B | access: full metadata, mixed targets | Curated route to the otherwise hard-to-find Barron's, Business Week, Forbes, Fortune, and HBS Bulletin profiles.
• ★★ | George F. Doriot Dies at 87; Molder of U.S. Businessmen | New York Times, 1987-06-03 | obituary | reliability: B | access: partial/paywalled | Contemporary life chronology and external assessment.
• ★★ | Milestones: Georges Frederic Doriot | TIME, 1987-06-15 | obituary | reliability: B | access: full | Concise contemporary verification of death, career, and teaching reputation.
• ★★ | The Talented Georges Doriot | Harvard Gazette, 2015-02-24 | retrospective | reliability: B | access: full | Harvard retrospective on biography and legacy; useful bridge to the Ante biography.
• ★★ | Kenneth Olsen: Venture Capital's First Entrepreneurial Hero | Institutional Investor, 2011-02-09 | retrospective with recollections | reliability: B | access: full | Deal narrative with Olsen recollections, ARD annual-report language, board role, milestones, and mentor framing.
• ★★ | Problems Facing Venture Capital Today Are Not New | Institutional Investor, 2010-09-22 | retrospective/excerpts | reliability: B | access: full | Processed for four attributed 1971 annual-report passages; verify against the original before expanding quotations.
• ★ | Has Venture Capital Strayed From Its Roots? | INSEAD Knowledge, 2020-07-06 | retrospective essay | reliability: B | access: full | Modern critique contrasts Doriot's development mission with financialized VC.

Coverage assessment

• Record shape: unusually strong archival/written record but weak publicly accessible direct audio/video. Baker, the Library of Congress, and the Computer History Museum hold the decisive primary evidence; much of it requires onsite use or a digitization request. The strongest voice targets are Harvard's 1 March 1966 class reel and CHM's 1979 birthday recordings.
• Name variants searched: “Georges F. Doriot,” “Georges Frederic Doriot,” “Georges Frédéric Doriot,” “General Doriot,” “George Doriot,” ARD, ARDC, and “American Research and Development.”
• Thin areas: no complete online list of prewar board seats; no comprehensive digitized ARD annual-report run; no publicly streamable authenticated Doriot recording; exact DEC ownership and realized-return figures conflict across HBS, CHM, Gompers, Hsu/Kenney, SEC records, and later retellings.
• DEC chronology correction: the SEC records DEC's registration effective on 18 August 1966. Its August 1968 record is a later ARD secondary sale, not the IPO.
• Stake and financing caution: credible sources report ARD's initial stake as about 70%, 77%, 78%, or 80%, depending on share basis and later adjustments. Preserve the distinction between the reported $70,000 equity purchase, a separately reported $30,000 loan, and later financing.
• Return-label discipline: roughly $38.5 million at the 1966 IPO is a reported market value, not cash proceeds; roughly $355 million around the 1971 distribution is a later reported value/gain. “500x,” “5,000x,” and “5,071x” are gross comparisons against $70,000, not audited fund net multiples. The reported 14.7% annual return with DEC and 7.4% without it for 1946–1971 comes through Liles/Doriot and later scholarship, not a public audited return series.
• Three best next leads: (1) request the Baker Doriot and ARD annual-report boxes plus the 1988 DVD; (2) use the Library of Congress container list to target lectures/diaries on risk, leadership, and R&D; (3) reconstruct DEC capitalization and ARD's liquidation from the 1957 proposal, ARD annual reports, DEC board records, and Hsu/Kenney's dataset.
• SCOUT saturation: source categories swept across direct writings, Harvard/LOC/Army archives, books, talks/podcasts/video, contemporary press, company records, academic studies, and deal evidence. Later phases should add exact reading-room citations rather than more low-grade web summaries.