Entrepreneurship / Venture Building

The structural pull at the field level is Creation in its most consequential form: not a painting or a piece of code but a whole working entity — a product, a team, a set of customers, a small economy — brought into existence where there was nothing before. Whatever a particular venture role looks like day to day, the field exists because someone decided that an idea was worth turning into a thing that employs people, serves customers, and survives on its own. The founder's deepest satisfaction and deepest terror come from the same fact: the thing exists because they made it exist, and it will stop existing if they stop.

Creation is the pull, but a venture is not one act of creation — it is a chain of gradients that a young company has to climb in order. First you build a product out of nothing (Creation). Then you have to get it in front of people and grow past your first handful of users (Spread). Then, somewhere around the point where it is working, the company tips into chaos — too many customers, too many half-finished processes, too many people who do not know who decides what — and the work becomes imposing order on the mess you created (Organization). All the way through, two other gradients run underneath: the relentless Resolution of things that are broken, half-built, or on fire, and the constant Judgement of which of a thousand possible things to actually do, because a startup dies far more often from doing too many things badly than from doing one thing too slowly. Understanding that the field is this whole sequence, not just the romantic first act, is the single most useful thing a 17-year-old can take from it.

The field is unusually broad in who it needs, and "entrepreneur" is not one job. A modern venture is built by founders who carry the vision and the risk, technical co-founders who build the actual product, growth people who figure out how to reach customers, operators who keep the expanding machine from tearing itself apart, and the investors who fund and advise the whole thing. These are genuinely different working lives that happen to share an ecosystem, and a person drawn to one may be poorly suited to another. The roles below reflect that spread on purpose, because the most common mistake a young person makes about this field is assuming that "starting a company" means being the charismatic founder pitching on stage, when the ventures that actually work are built by a small number of very different people pulling in the same direction.

Now the honest part, because entrepreneurship is romanticised more aggressively than almost any field a teenager will encounter, and the honesty is the most useful gift here. The base rate is failure. In the UK only around 42% of new businesses survive to five years, and close to two-thirds fail within the first three [from salary and job-posting data; PwC and ONS-derived business-survival data, 2025]. Venture capital — the version of entrepreneurship that gets written about — is statistically almost irrelevant to the average founder: only a tiny fraction of startups ever raise it, and even among those that do, most never return their investors' money [from salary and job-posting data; industry analyses, 2025]. The money is not where outsiders think it is, either. A UK founder who raises a seed round typically pays themselves a modest salary, often £40,000–£80,000 and sometimes nothing at all in the earliest months, while taking on far more risk and stress than a salaried job would carry [from salary and job-posting data; SeedLegals UK founder-salary data, 2025]. The compensation that makes the risk rational is equity — a share of a company that is, statistically, most likely to be worth nothing. People who build durable careers here are the ones who would do it anyway, because they cannot tolerate not building.

Entry has no gate, and that is the field's great democratic feature. There is no licence, no required degree, no exam. You become an entrepreneur the moment you start building something and trying to get people to pay for it, which a 17-year-old can do this weekend with a free website and an idea. What exists instead of credentials is a support ecosystem worth knowing about: accelerators such as Y Combinator (whose standard deal in 2025 is around $500,000 for roughly 7% of the company, plus mentorship and a network) and European programmes like Seedcamp and Entrepreneur First take very early founders and compress years of learning into months, though acceptance rates run as low as 1–3% [from salary and job-posting data; accelerator program terms and Y Combinator published deal, 2025]. For those who want to build across borders, structured routes exist — the UK's Innovator Founder visa requires a business that an authorised endorsing body certifies as innovative, viable, and scalable [official regulator; UK Home Office / GOV.UK, 2025], and Portugal's Startup Visa runs through a network of nearly a hundred IAPMEI-certified incubators that host and mentor founding teams of up to five [official; IAPMEI / Startup Portugal, 2025]. But the most persuasive credential in the entire field is the same one games and design reward: something you actually built and put in front of real people. Founders who have shipped a thing — even a tiny, failed thing — are taken seriously in a way that founders with only a plan never are.

People and jobs. The field has no employer and therefore no conventional supply-demand equilibrium — but its intake profile is restructuring on every measurable axis. Q1 2026 seed dollars +31% to ~$12B while deal count -30% to ~3,800 [happening now; Crunchbase/PitchBook-derived]. AI took ~86% of US venture dollars in H1 2026, $355.9B of $412.7B [happening now; PitchBook via SiliconANGLE]. OpenAI and Anthropic alone took ~43% of all H1 2026 startup funding [happening now; Crunchbase]. YC batches are ~60% AI companies vs ~40% in 2024 [reported by people in the field]. Composition is shifting hard even where aggregate dollars rise.

The main limit on getting in. There is no credential gate — the Mode 1 narrative is emphatic that entry is ungated and that is the field's democratic feature. The binding constraint is capital, and it has tightened rather than loosened: bigger cheques into ~30% fewer seed companies. Note this constraint binds only the venture-funded path; the bootstrapped path's binding constraint has genuinely loosened, which is what makes the field's key finding.

Entry-level jobs. Contracting for the salaried routes into the field (early startup employee, junior VC). Expanding for the unsalaried route (founding directly). The single enum value cannot carry this split; see the asymmetric entry level impact severity segments. Recorded as 'contracting' because the enum should describe *demand for labour*, and founding is not demand for labour — it is self-employment.

Cost of training. No required degree or credential means no field-specific educational debt gate. Recorded low rather than doesn't apply here because founder compensation is genuinely poor — UK seed-stage founders typically pay themselves ~£40,000-£80,000 and sometimes nothing in the earliest months [from salary and job-posting data; SeedLegals UK founder-salary data, 2025, via Mode 1] — so any existing debt service is a real and underdiscussed constraint on who can afford to found. The gate is not credential debt; it is the ability to survive on little.

Nothing about this field's AI story is softened by a workforce shortage — there is no shortage of people who want to found companies, and no employer bidding for them. The AI story here must be read against a base rate of failure (only ~42% of UK new businesses survive to five years) rather than against scarcity. AI lowers the cost of attempting and does not improve the odds of succeeding, and those are different things that the field's promotional culture routinely conflates.