Frequently asked questions

What is The Slingshot?

The Slingshot is a free, open-source game where you found and run an AI startup in the UK. You play quarter by quarter over four years, racing rival firms to reach three milestones before the money or the time runs out. It runs in the browser, so there is nothing to install and no account to create.

Is it really free?

Yes, completely. There is no premium tier, no account and no catch. The game is an open, not-for-profit educational project.

How long does a session take?

A full run takes about twenty-five to forty minutes, and returning players tend to be quicker.

Is it fun?

Yes. It was designed by a gamer for gamers. It is a game first and an educational simulation second. Many commercial educational simulations are worthy and dull, lacking the humour and engagement of commercial games. It is also hard, like many commercial games. The difficulty is part of the challenge. It is designed for replayability. Some players have gone back thirty times chasing a better exit. As one Master's student put it, “it was a fun and motivating way to experience how business decisions are made under pressure.”

Where should I start?

Start in Seed mode with any venture that catches your eye, or build your own venture with Ada. Use your first run to learn the ropes rather than to win, and read the journey report at the end. Once you have found your feet, move to Core mode, where the competition is live. The how-to-play page explains the systems.

How hard is it?

It is challenging, but not impossible. Most first runs end with the money gone or a milestone missed, and the game does not soften the landing. Seed mode gives you more money and attention while you learn; Core is the standard game; Extreme removes the guide rails. As you play, you will get better at using the systems of the game, and learn when and how to raise funding, allocate attention, pivot, and deal with funders. The game is designed to improve your judgement under uncertainty in the entrepreneurial process.

What should I focus on?

Focus on the next milestone. Each milestone asks for particular capabilities by a deadline, so read the requirements early and point your attention at them. You also need to keep an eye on your cash, because raising money takes time and the game ends in the quarter you cannot pay your bills. Markets, hiring and events all feed into these two things.

What decisions will I face?

You will face the same decisions founders face: whether to publish your research or protect it, which market to back and when to pivot, whose money to take and what it costs you in equity, who to hire on a payroll you can barely cover, and what to do when your AI does something you did not ask for. There are nearly six hundred decision situations in the pool, drawn from real startup experience, and no run sees more than a fraction of them.

Why is the game set in the UK?

The UK has a dynamic and innovative AI ecosystem. It is home to over 6,000 AI startups, world-leading universities and a thick market of angel and venture capital investors. The game is also designed in “Peaky Blinders mode”, using the rich context of the UK to enrich and enliven the experience. You do not need to know anything about the UK to play, as all the institutions and organisations are explained in the text, and you can always ask the player assistant.

Do I need to know about AI?

No. The game is written in a non-technical way, and all the decisions are explained in plain English. There are also help buttons to explain the game's systems, such as the token system, and the player assistant can answer questions as you play.

What will I learn?

You will pick up the terms and concepts of technology venturing by using them: runway, dilution, term sheets, pivots and milestones. The game is designed to help you build confidence and understanding of how ventures work, and to give you practice before plunging in.

What mistakes do players often make?

Players tend to make the same mistakes. The most common is doing whatever the funders suggest instead of thinking for yourself. Many players take the first offer on everything, when staff, funders and partners all expect a negotiation. Some never hire, so the founding team ends up carrying the whole company, while others spread their attention everywhere and commit to nothing. Grant funding is often ignored, even though it is money that does not cost you equity, and many players never open Market Opportunities, so a pivot costs them full price. Finally, players often pick a venture, location and milestone path that do not match their ambition. The game rewards deciding what you want and setting the company up to achieve it.

What happens when my company fails?

You get a journey report, which is a PDF case study of your run: what you chose, what it cost and how it ended. Read it, take it to Schumpeter for a debrief, and then play again with a different founder, venture or plan. A question worth asking between runs is how much of the result was judgement and how much was luck.

Do I play against others?

Yes, in two ways. Inside every run, rival firms race you to the same milestones, raise money and take your customers. Between runs, Trajectory in Core mode benchmarks your result against everyone who has recorded one, and a class can play a shared challenge from the same starting position.

Can I get help?

Yes. The player assistant is a free tool for players. You can ask it how a mechanic works, what a term means, or how to think about your next move, and it answers in plain language based on the game's own materials.

Can't I just get an AI to play it for me?

You can try. In our experience, AI models are not very good at the game: they lose track of their own cash position, struggle to plan across sixteen quarters, and cannot play the mini-games at all. More to the point, there is nothing to win by it. There are no grades or prizes attached to playing, so an AI playing for you is just you not playing. The game is designed to improve your judgement, and that only works if you are the one making the decisions. If you do build an AI that plays it well, we would like to hear about it.

Why is every run different?

No two runs are alike. The game draws from a pool of nearly six hundred decision events, and a single run only ever sees a fraction of them, so each run deals you different situations in a different order. The rivals behave differently each time, funding rounds unfold differently, and the same choice can turn out differently, because outcomes involve chance as well as judgement. On top of that there are twelve built-in ventures, each with its own locations and markets, different founder profiles and three modes, and with Ada you can bring a venture of your own. This is also why there is no strategy guide that always works, and why an AI cannot simply play out a winning path. You can go back thirty times and still be surprised.

Can I suggest features or improvements?

Yes, please. The Slingshot is a community project, and players' ideas have shaped every version so far. If something confused you, broke, or ought to exist, email ammon.salter@wbs.ac.uk. Your experience and ideas can shape the next version of the game.

What are the mini-games?

There are three moments in the game that you play rather than choose from a menu: working your booth at a conference, delivering a thirty-second pitch to investors in London, and closing deals under time pressure at demo day in Manchester.

What is Ada?

Ada builds a venture from your own idea. In a short conversation you settle the idea, the market, a company name and a location, and Ada saves a small file that the game loads. You then play your own startup rather than one of the twelve built-in ventures. Meet Ada.

What is the Schumpeter debrief?

Every run ends with a journey report recording your venture's history. You can upload it to Schumpeter and talk the run through with him: what you chose, what it cost, and what you would do differently.

Are you harvesting my data?

No, we are not. The game lives on your computer, not a server. Your game file is yours. We only use trackers to assess general use by country and time. We have no way to identify you. We recommend using nicknames, not your real name. No one has any record of your AI startup idea but you. Ada does not record your conversation.

Can you steal my AI startup idea?

No, we can't. There is no record of your idea. Your idea stays yours. The game is designed to support you, not steal from you.

Where do I go from here?

An educator once described the game as 'dry swimming', and they were right. But dryland training is part of how swimmers get fast: Michael Phelps did plenty of it, and Summer McIntosh lifts weights, kettlebells and core work. Like swimmers, entrepreneurs can learn a great deal before they dive in. The game is designed to help you test out ideas, and to gain the confidence and understanding about how to take your idea forward. To go further, try to build out your idea and then apply to accelerators, incubators, angels, funders and venture competitions to take it forward. Use the resources of your university or local area to help you, such as mentoring services, student venture clubs, training programmes and classes.

Who built it?

The Slingshot was created by Ammon Salter (Warwick Business School), with Stefano Baruffaldi (Politecnico di Milano) and Federico Bignone (Warwick Business School), using AI-assisted development tools rather than a software studio. It is an open, not-for-profit educational project, built and maintained by a small team alongside their other work, and licensed CC BY-NC-SA 4.0.

What do students need?

A modern web browser. Nothing to install, nothing to pay, no accounts to create. The game sets no cookies and holds no student data; the only things that leave the browser are anonymous, aggregate play counts and an optional, opt-in leaderboard.

What do students learn?

Generative AI and its use in entrepreneurial settings; entrepreneurial strategy in a nascent industry, where markets, categories and rules are still forming and there is rarely a single right answer; strategic decision-making under uncertainty; entrepreneurial finance, including runway, dilution, term sheets and non-dilutive grants; technology venturing and commercialisation, turning advanced technology into a business; university-industry interaction, engaging the science base that sits upstream of a deep-tech venture, deciding whether to publish or protect, and building university partnerships; managing investors, regulators, customers and partners; and learning through failure and structured reflection.

Do I need to train my students to play?

No. You post a link. The game teaches itself: guided play for first runs, a tutorial, help buttons throughout, and the players' AI assistant for anyone who wants more. There is nothing to install and nothing to set up. Students can replay it as often as they like, in and out of class. Some play once, others play thirty times. The game has optional depth for those who want to invest time in it, but it is still playable for those happy to try it once.

How much further you go is up to you. The game works on its own from a single link. The Educators Pack is there if you want to tie it into your course: session plans, classroom exercises, mini-cases to read alongside runs, and class reports on how your students played. Because the game is free, the use case is shaped by your needs rather than a commercial model. There are no seats, no coursepacks and no per-student charges. Some colleagues spend ten minutes on it; some build whole sessions around it.

Do I need to master the game to use it?

No. Many faculty have used the game successfully with their students without ever playing it as a game themselves. You do not need to be a gamer (or even like games yourself!) to use it in your class. The game teaches concepts and ideas that most educators understand well already. The Educators Pack was created to help you with the resources you need to lower the challenges of adoption. You also have a secret weapon: the AI Educators Assistant, which can answer questions before class and in real time as they come up. It is remarkably good, and it gives you insights into the game that are not available to the players. It will save you, if you let it. Some of your students will end up knowing the game far better than you. Gamers will look for hacks to gain advantage, just like entrepreneurs do. That is part of the learning experience. The more they play, the more they will learn.

Is it accessible?

Yes. The game is developed against WCAG 2.2 AA. It can be played end to end by keyboard and works with screen readers. If you hit a barrier, email ammon.salter@wbs.ac.uk and we will fix it.

What courses and levels does it suit?

Entrepreneurship, entrepreneurial strategy, innovation management, technology venturing and technology commercialisation, at undergraduate, postgraduate and executive levels. The three modes (Seed, Core, Extreme) let you match difficulty to the cohort.

If it is so good, why is it free?

Because it was built by academics, not a company. The Slingshot came out of an ESRC-funded research project on AI ventures, and giving it away is part of the point: we want as many people as possible to learn from it. The development team gains credit through citations rather than sales, which is how academic work usually works. The game is free for education and research, and commercial users pay for a licence, which helps keep it that way. There is no premium version waiting behind the free one. What you play is the whole game.

How does it compare with other entrepreneurship simulations?

Commercial simulations are typically licensed per student, need accounts and setup, and are often built around US examples and longer timetables. The Slingshot is free and open source, runs in the browser in twenty-five to forty minutes with no setup and no student accounts, is grounded in published research on AI ventures and the UK ecosystem, and comes with a free Educators Pack. Because the licence is open, you can adapt the game to your course rather than fitting your course to the game. Short free runs also make it replayable — many students play repeatedly, unprompted — and open access gives it uses a licensed product cannot match: a link shared for self-directed play, homework before class, an in-class tournament, or reflective assessment. Educators should judge any simulation against their cohort and learning objectives; those attributes are what The Slingshot offers.

Why the UK?

The UK has a dynamic and innovative AI ecosystem, and the game was designed in “Peaky Blinders mode” to give it a feel for the local context and institutions. It covers all the regions and countries of the UK, and the choices available to players reflect the characteristics of place. Entrepreneurship does not take place on the head of a pin. Plenty of cultural and educational products are set in the UK and are loved by people who have never been here. You can love the Arctic Monkeys and never visit Sheffield. Many commercial educational simulations are set in the US, which is often an ill-fitting context for students elsewhere.

Is the game still being developed?

Yes, continuously. Each version adds features: Version 2 brought Seed mode, competitive play through Trajectory, the Market Opportunity Navigator and token spending; Version 2.1 added Ada, four new sectors, the Schumpeter debrief, rebuilt mini-games, accessibility throughout and time-stamped play for assessment. The aim is a free resource that rivals anything in the commercial market, and is more fun to play.

Is it based on research?

Yes, in both its theory and its content. The game builds on the attention-based view of the firm (Ocasio, 1997): attention is the scarce resource, and the game's central tension is allocating it between competing goals, building internal capability while engaging the outside world that holds the resources a venture needs. That tension is also the question of the ESRC-funded project behind the game, Profiting from Science (ES/X003949/1), which studies how AI ventures form, grow and engage the science base. Its market-entry system embeds the Where to Play framework (Gruber and Tal, 2017), and its 590+ events draw on the literature on technology venturing.

Can I use it for assessment?

Yes, through reflection rather than results. Game outcomes make a poor measure of learning, because the game has a strong degree of randomness by design and a result reflects luck as well as judgement. What the design supports is a reflective essay: every run ends with a report, a PDF the student keeps, recording the venture's history, and students write about the choices they made during the playthrough, why they made them, and what they would do differently. That reflection is naturally the student's own. Only the student knows why they chose as they did, and the game's detailed mechanics are deliberately hidden, so the insight in a good essay comes from having played and thought, not from anything that can be looked up. The Educators Pack includes assignment ideas built exactly this way. Every report also carries a time-stamped frank: a unique run code, the time played and the pace of decisions, so a submitted report can be checked as the student's own run.

Can students play in teams?

Yes. A pair of students can share one venture and make every decision jointly, which turns each quarter into a strategic argument, over hiring, funding, markets and pivots, of the kind a real founding team has. A class can also run a shared challenge: everyone starts from the same conditions and races to build the next AI unicorn, the game's ultimate prize, with a shared join code making the cohort its own anonymous leaderboard.

Can students play competitively?

Yes, in three senses. Inside the game every venture races three named rival firms in its sector. Between players, Core mode includes live competitive benchmarking, so students can compare their run against their class and against everyone who has played, and educators can run zero-cost in-class tournaments with no registration and no personal data captured. And players can compare themselves against themselves: benchmarking across their own earlier runs makes improvement visible from one replay to the next. Seed switches benchmarking off so beginners can learn without it.

How do I cite it?

Salter, A., Baruffaldi, S., & Bignone, F. (2026). The Slingshot: An AI Startup Simulation. Zenodo. https://doi.org/10.5281/zenodo.20749017

How do I get the Educators Pack?

Email ammon.salter@wbs.ac.uk. It is free. The pack includes an educator AI assistant: ask it how to fit the game to your course, module or assignment, and it answers from the game's own materials.

What is the Trajectory Wizard?

A class-reporting tool in the pack. Pool your class's end-of-game reports, attach the Wizard's instruction file to an AI assistant, and get back a plain slide deck comparing the class, with draft questions for the debrief. Every number traces back to the reports.

Can I pair the game with real company examples?

Yes. The pack includes short mini-cases of real UK AI startups — Darktrace, Sylvera, Robin AI, Multiverse, Synthesia, Lendable, Wayve and Isomorphic Labs — each pairing with a sector, so students play the decisions and then read how the real company faced the same ground. The cases also work the other way round: read before a run, they give students a feel for the sector they are about to take on.