# The New Sweden – 15 Points for the Future
**Core thesis: The problem isn't people, it's friction.**
**Core rule: AI where it reduces friction, humans where they exercise power.**
*Joakim Jardenberg* | [https://15p.jardenberg.se](https://15p.jardenberg.se)
---
## About this project
It all started with a Facebook post by Troed Troedson. He challenged: "Parliamentary parties! Hold a press conference each with 15 points that ease people's journey instead of the opposite. But hold a workshop first so it doesn't turn out as clumsy as it usually does."
Said and done. I workshopped with my team and came up with 15 points that are, of course, very much my own. Leaning heavily on the internet, digitalisation in general and AI in particular. And which are, equally obviously, a work in progress. Better together, as always.
### How we built the site
From walk-and-talk to finished site — this project is a collaboration between human and AI.
It started with a voice recording. I went for a walk and talked through my thoughts about Troed's post, about integration, about AI and friction. Manus — my AI agent — transcribed the recording, extracted the 15 points and structured them into a manifesto with headings, key figures, quotes and body text.
We worked iteratively. I gave direction and feedback, Manus built and suggested. First came the content — the 15 points with their investment figures and arguments. Then the design: we landed on "Nordic Monumentalism" with a warm cream background, ochre accents, Playfair Display typography and a sense of gravity without being heavy.
Manus did its own research to give the points credibility — figures from the OECD on AI and language training, data on validation times for foreign qualifications, Global Innovation Index rankings, and statistics on shortage occupations in Sweden.
The 15 illustrations were created with Nano Banana — an AI image generation tool — in a warm 1950s editorial style. I uploaded reference images with a vibe I liked, and Manus generated 15 unique images matching each point's theme.
Along the way, we created skills — reusable knowledge packages — that document how we did things. This means that next time I or anyone else wants to build something similar, it will be faster and better.
On the technical side, Manus built the site with React, Tailwind, tRPC and Express — with server-side rendering of OG tags so the right image and title appear when someone shares a point on social media. Comments and voting with a database behind it. Hosting via Manus with a CNAME pointer to my own domain.
Read more: [Our full write-up on how you build something like this in 2026](https://ai.jardenberg.se/15p-build)
---
## Contents
1. [One Sweden In, Not Fifteen Queues](#punkt-1)
2. [100-Day Competence Guarantee](#punkt-2)
3. [Swedish Through Work, Not Before Work](#punkt-3)
4. [National Fast Track to Shortage Professions](#punkt-4)
5. [First Real Job Within 180 Days](#punkt-5)
6. [Settle by Opportunity, Not by Passivity](#punkt-6)
7. [Children First, Always](#punkt-7)
8. [Hard on Exploitation, Soft on People](#punkt-8)
9. [Zero Tolerance for Discrimination](#punkt-9)
10. [Digital Membership from Day One](#punkt-10)
11. [Europe's Easiest Country to Start a Business In](#punkt-11)
12. [Student to Builder of Sweden](#punkt-12)
13. [Circular Mobility as Strength](#punkt-13)
14. [Fast and Predictable Rule of Law](#punkt-14)
15. [Integration as a National Elite Discipline](#punkt-15)
---
## Point 1: One Sweden In, Not Fifteen Queues
**A single digital path in – with a human guide and an AI guide in your pocket**
*Platform & Onboarding*
We're building a single path into Sweden: a digital onboarding experience, not an administrative obstacle course. Every person should encounter a unified interface where identity, documents, language support, health contact, school track, work track, and civic information are connected. There will be both a human guide and an AI guide in your pocket: AI translates, explains, summarizes, and warns when something is missing, while the caseworker makes assessments and takes responsibility. The state should function more like a smart platform and less like fifteen government islands playing ping-pong with people's lives. Today, people are sent between the Migration Agency, the Employment Service, the Tax Agency, the municipality, and the Social Insurance Agency as if that were a reasonable system. It isn't. It's a design problem. Digg has already approved e-identification for people without Swedish personal numbers and expanded identity matching – the tools exist, what's missing is will, pace, and design.
> "What today looks like an integration problem is often just a design problem the state hasn't taken seriously yet."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 18 bn |
| People reached per year | 200,000+ |
| Agencies connected | 15+ |
| Processing time (target) | -70% |
| Languages in AI guide | 40+ |
| Full rollout | 24 months |
---
## Point 2: 100-Day Competence Guarantee
**A digital competence passport that stops starting with suspicion**
*Competence & Work*
We're introducing a digital competence passport. Degrees, certificates, work samples, references, and experience are uploaded once, read by machine, compared against Swedish requirements, and result in a clear answer: approved, supplement this, or here's an alternative track. AI is used to interpret documents, match against qualification requirements, and suggest supplements. The human makes the decision, but the system must stop starting with suspicion. A piece of paper that doesn't fit a Swedish folder shouldn't erase an entire working life. Today, validation of foreign qualifications takes an average of 18 months. During that time, surgeons drive taxis and engineers clean stairwells – not because driving a taxi or cleaning is wrong, but because we're wasting competence we desperately need. It's not just unfair, it's economic insanity. Sweden ranks second in the Global Innovation Index – imagine what we could do if we actually used all the competence that's already here.
> "A piece of paper that doesn't fit a Swedish folder shouldn't erase an entire working life."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 15 bn |
| Validations per year | 90,000 |
| Max time to decision | 100 days |
| Societal savings | SEK 40 bn/year |
| Regional validation centres | 50 |
| Industries in system | 120+ |
---
## Point 3: Swedish Through Work, Not Before Work
**A personal AI language coach in your pocket – connected to your job, your education, your everyday life**
*Language & Learning*
Everyone who needs it gets Swedish in their pocket: a personal language coach with speech training, occupational vocabulary, dialogue simulations, real-time translation, and feedback connected to the specific job, education, or everyday life they're in. We shouldn't have a model where people first wait for years to become sufficiently Swedish before they start contributing. Language is best learned when it's used, and AI makes it possible to practice a little, often, relevantly, and at your own level. This is perhaps the most underestimated lever of all. The OECD highlights AI as particularly promising for language training in integration. A Somali-speaking doctor can save lives in Somali while learning Swedish. An Arabic-speaking engineer can solve technical problems in Arabic from the first week. Parallel competence and parallel language – not sequential. SFI in its current form is a waiting hall, not a runway. We're rebuilding it from the ground up.
> "Language is best learned when it's used. Not in a classroom you waited a year to enter."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 22 bn |
| People reached per year | 180,000 |
| Languages in AI coach | 40+ |
| Expected GDP effect | +0.8% annually |
| Career tracks with parallel language | 80+ |
| Full rollout | 18 months |
---
## Point 4: National Fast Track to Shortage Professions
**Digital fast tracks with validation, simulations, and AI-supported knowledge tests**
*Competence & Work*
We're building digital fast tracks to shortage professions with validation, simulations, AI-supported knowledge tests, supervised practice, and paid supplementation. A nurse shouldn't have to start from zero because the state lacks imagination, and an electrician shouldn't get stuck between PDFs and binders. Technology should be used to compress the path to Swedish licensure, not to create more waiting rooms. When someone knows the job, the system should help prove it faster. Sweden currently has staff shortages in over 100 occupational categories – from assistant nurses to civil engineers, from welders to specialist doctors. Meanwhile, we have tens of thousands of people with exactly that competence stuck in a system designed to slow down. It's like having a highway with a speed limit of 30. We're removing the unnecessary speed limits.
> "Technology should compress the path to Swedish licensure, not create more waiting rooms."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 12 bn |
| Fast tracks in shortage professions | 100+ |
| Participants per year | 45,000 |
| Time to licensure (target) | -60% |
| Societal savings | SEK 25 bn/year |
| Internship placements | 30,000/year |
---
## Point 5: First Real Job Within 180 Days
**From CV theater to competence proof – AI matching that sees potential, not templates**
*Competence & Work*
We're moving from CV theater to competence proof. AI-supported matching can connect people with employers based on skills, language progression, work samples, references, and actual potential – not just Swedish templates for cover letters. Every person gets a combination of a digital coach and a human job matcher. AI should replace waiting, not people. We must stop pretending that CV writing is integration. Today, it takes an average of over two years for a newly arrived person to get their first real job. Two years of waiting, uncertainty, and competence loss. It's not just an individual problem – it's a system failure. Employers are screaming for competence while competence sits in waiting rooms. We're building the bridge between them with technology that sees what people can do, not just what they've written on a piece of paper.
> "We must stop pretending that CV writing is integration."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 10 bn |
| Matches per year | 120,000 |
| Target: time to first job | 180 days |
| Employers in network | 50,000+ |
| Socioeconomic gain | SEK 35 bn/year |
| Languages in matching system | 40+ |
---
## Point 6: Settle by Opportunity, Not by Passivity
**Data showing where the jobs, schools, and housing are – not control, but real choices**
*Platform & Onboarding*
We use data to see where the jobs are, where schools work, where housing exists, and where the path to self-sufficiency is shortest. Not to control people like chess pieces, but to give them better information and real choices. A national tool should be able to show: here are jobs in your field, here's language training, here's reasonable commuting, here are housing openings. It's strange that we accept more intelligence in a travel app than in the state's decisions about people's futures. Today, people are placed in municipalities with high unemployment, poor public transport, and few opportunities – not because it's best for them, but because it's easiest for the system. It's like giving someone a map without roads. We're building the map with roads, real-time data, and personal recommendations. Every municipality that receives newcomers should be able to show exactly what opportunities exist – and be measured on how well they deliver.
> "It's strange that we accept more intelligence in a travel app than in the state's decisions about people's futures."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 8 bn |
| Municipalities with real-time data | All 290 |
| Data points per municipality | 200+ |
| People reached per year | 80,000 |
| Time to self-sufficiency (target) | -40% |
| Open data sources | 50+ |
---
## Point 7: Children First, Always
**Digitally equipped schools for multilingualism – technology supports, the teacher leads**
*Children & Education*
Every newly arrived child should meet a school that is digitally equipped for multilingualism. Recorded lessons, translated guardian flows, AI-supported language support, digital study guidance, adaptive exercises, and simple ways for home and school to understand each other should be standard, not project funding. But the principle is crystal clear: technology supports, the teacher leads. We shouldn't build bot schools. We should give good teachers much greater reach. The National Agency for Education emphasizes clear frameworks and teacher responsibility for AI in schools – and that's exactly right. But the frameworks should enable, not restrict. A child who speaks Dari at home and Swedish at school doesn't have one language too many – they have one more language than most. Every school with more than 20% multilingual students gets double resources, not fewer. We recruit 15,000 multilingual educators and pay them as specialists. The OECD shows that early educational interventions yield the highest return of all societal investments.
> "We shouldn't build bot schools. We should give good teachers much greater reach."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 35 bn |
| Children reached | 350,000 |
| New multilingual educators | 15,000 |
| Return per krona | 7:1 (OECD) |
| Languages in school platform | 60+ |
| Full rollout | 3 years |
---
## Point 8: Hard on Exploitation, Soft on People
**Data analysis that exposes exploitation – and a digital safe room in the worker's pocket**
*Rule of Law & Protection*
We use data analysis and AI to detect sham employment, unreasonable wage levels, fraudulent companies, exploitation in subcontractor chains, and suspected patterns of labor crime far earlier than today. At the same time, the worker gets a digital safe room in their pocket: contracts in their language, wage checks, rights guide, simple reporting, anonymous whistleblowing, and direct contact to support. Technology shouldn't just help the state control. It should help people understand their rights and dare to use them. Labor crime costs society an estimated 100 billion kronor annually in lost tax revenue, distorted competition, and human suffering. Those hit hardest are those with the least power to defend themselves. It's a perverse irony that the same system that makes it hard to get a work permit makes it easy to be exploited. We're reversing that: easy to work legally, impossible to exploit someone.
> "Technology shouldn't just help the state control. It should help people understand their rights and dare to use them."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 8 bn |
| Workers with digital protection | 500,000+ |
| Expected savings | SEK 30 bn/year |
| Languages in safe room | 40+ |
| Inspections with AI support | ×5 more |
| Implementation time | 18 months |
---
## Point 9: Zero Tolerance for Discrimination
**Anonymized screening, skills tests before gut feeling, and AI systems audited for bias**
*Rule of Law & Protection*
We're confronting both analog and digital discrimination. Public recruitment must use anonymized first screening, skills tests before gut feeling, and ongoing transparency in outcomes. AI systems used for matching, selection, or support – especially in jobs and education – must be auditable, explainable, and possible to shut down if they reproduce old prejudices. It's not enough to digitize injustice. Sweden should become the country that measures bias as seriously as we measure budgets. The EU AI Act already applies in stages and gives us a framework – but we'll go further. The Equality Ombudsman gets quadrupled resources and the right to impose significant fines. Multilingualism becomes a formal merit in all public services. Speaking Tigrinya and Swedish isn't just as good as speaking only Swedish – it's better. The goal: within five years, the employment gap between native and foreign-born should be gone.
> "It's not enough to digitize injustice. Sweden should become the country that measures bias as seriously as we measure budgets."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 6 bn |
| Employers affected | 85,000 |
| Equality Ombudsman new budget | ×4 |
| Target: employment gap | 0% (5 years) |
| AI systems under audit | All public |
| Implementation time | 2 years |
---
## Point 10: Digital Membership from Day One
**Secure digital identity, government mail, and a clear next-step function on your phone**
*Digital Infrastructure*
You shouldn't have to become analog in a digital country just because you just arrived. From day one, people should be able to get secure digital identity, government mail, a consolidated case overview, bookings, language support, civic information, and a clear next-step function on their phone. This isn't decoration. It's infrastructure. What today looks like a personal number problem is often just a design problem the state hasn't taken seriously yet. Digg has already expanded e-identification for people without Swedish personal numbers and more login options in public e-services. The groundwork is done. Now we need to connect it into a coherent experience. Sweden ranks among the world's most digitalized countries – but only if you already have a personal number, a BankID, and know how the Tax Agency works. For everyone else, Sweden is still analog, bureaucratic, and incomprehensible. We're fixing that.
> "What today looks like a personal number problem is often just a design problem the state hasn't taken seriously yet."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 12 bn |
| People reached from day one | 200,000+/year |
| Digital services in platform | 50+ |
| Languages in interface | 40+ |
| Time to digital identity | 24 hours |
| Digital export potential | SEK 20 bn/year |
---
## Point 11: Europe's Easiest Country to Start a Business In
**From idea to registered company, VAT, bank, and first quote – on your phone, in your language**
*Business & Entrepreneurship*
We're building a multilingual, AI-supported business track where an idea can go from thought to registered company, VAT, bank, accounting support, first quote, and first public procurement without first having to become an expert in Swedish government language. Sweden has everything to gain from seeing the entrepreneur in those who come here, not just the future employee. Internet logic applies here too: lower the thresholds, open the interfaces, let more people build. Company building should fit in your phone, not first in a binder. Immigrants already start businesses at a higher rate than native Swedes – and they bring global networks, new markets, and business ideas we never would have thought of ourselves. In the US, immigrants have founded over half of all unicorn companies. We want the same force here. Every new entrepreneur gets an AI assistant for accounting and compliance, a human business mentor, and access to a New Start Fund with microloans and advisory services.
> "Internet logic applies here too: lower the thresholds, open the interfaces, let more people build."
### Key Metrics
| Metric | Value |
| --- | --- |
| Fund capital | SEK 30 bn |
| New companies per year | 25,000 |
| New jobs (5 years) | 150,000 |
| Tax revenue (5 years) | +SEK 45 bn |
| Time from idea to company | 1 day |
| Languages in business track | 40+ |
---
## Point 12: Student to Builder of Sweden
**Automatic digital transition track from graduation to employment, research, startup, or public service**
*Competence & Work*
Those who have spent years of their life studying here shouldn't be met with an administrative 'thanks and goodbye.' We need an automatic digital transition track from graduation to employment, research, startup, or public service. AI can help map competence, match against shortage professions, find co-founders, and connect people to regional innovation environments. The dumbest thing a knowledge country can do is educate global talent and then actively make it harder for them to stay and contribute. Sweden invests hundreds of millions in every international student's education – and then we do everything we can to push them out. It's like planting a tree, watering it for five years, and then digging it up. Instead, we're building a seamless transition: before the student graduates, the system has already mapped opportunities, connected contacts, and suggested next steps. Residence permits for work after graduation should be automatic, not a new bureaucratic hurdle.
> "The dumbest thing a knowledge country can do is educate global talent and then actively make it harder for them to stay."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 5 bn |
| International students/year | 40,000 |
| Target: stay after graduation | 70% (from 30%) |
| Automatic transition tracks | 4 (employment/research/startup/public) |
| Innovation environments connected | 50+ |
| Time to residence permit | Automatic |
---
## Point 13: Circular Mobility as Strength
**Portable credentials, digital competence wallets, and the diaspora as network – not loss**
*Mobility & Networks*
We must stop thinking that mobility only has value if it ends in permanent standstill. In a connected world, people can contribute to Sweden from Malmö, Mumbai, and Madrid at different stages of life. That's why we're building portable credentials, digital competence wallets, easy re-entries, and strong diaspora networks. The diaspora is not a loss. It's a network. Sweden should be a hub in people's lives, not an administrative needle's eye. The old model – come here, stay forever, or disappear – doesn't fit a connected world. Ireland built its tech boom partly through its diaspora. Israel has done the same. India likewise. Sweden has a global diaspora of hundreds of thousands of people with Swedish educations, Swedish networks, and Swedish values. We'll make it ridiculously easy to come back, contribute remotely, invest, research, teach, and collaborate – regardless of where in the world you are.
> "The diaspora is not a loss. It's a network. Sweden should be a hub in people's lives."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 4 bn |
| Swedish diaspora globally | 500,000+ |
| Digital competence wallets | 1 million |
| Re-entries per year (target) | 25,000 |
| Diaspora networks | 50 countries |
| Implementation time | 2 years |
---
## Point 14: Fast and Predictable Rule of Law
**Track your case like a package – with better law and greater dignity**
*Rule of Law & Protection*
Every person should be able to track their case like you track a package today, but with better law and greater dignity. Clear status, expected processing time, missing documents, next steps, responsible function, translation in your own language. AI can summarize, structure, prioritize, and flag – but the exercise of power must never be hidden behind a black box. Slow ambiguity is not neutral. It's a political choice. Today, people wait months, sometimes years, without knowing where in the process they are. They call authorities that don't answer. They send documents that disappear. They live in limbo while their lives tick away. That's not legal certainty – it's legal uncertainty. We're building a system where every case has a clear timeline, where AI flags bottlenecks before they arise, and where no person needs to call an authority to find out what's happening with their life. IMY's guidelines for AI in public administration give us the framework – now we're building the reality.
> "Slow ambiguity is not neutral. It's a political choice."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 10 bn |
| Cases with real-time status | All |
| Processing time (target) | -50% |
| Languages in case system | 40+ |
| People reached per year | 300,000+ |
| Appeals (target) | -60% |
---
## Point 15: Integration as a National Elite Discipline
**Open dashboard, open APIs, and open test beds – that's how the internet works, and how a modern country should work**
*Governance & Transparency*
We're building an open integration dashboard for the entire country. Municipality by municipality, agency by agency, in real-time where possible: time to first job, time to competence assessment, school progression, entrepreneurship, housing stability, participation, and trust. And we open data, APIs, and test beds so that civil society, researchers, startups, and municipalities can build better solutions on shared infrastructure. That's how the internet works. That's also how a modern country should work. Measure what works, stop winning debates with anecdotes. Every point in this program gets measurable goals and a responsible minister. Results are presented in parliament and on the open platform – accessible in 40 languages – where every citizen can follow developments in their municipality. If we fail, it should be visible. If we succeed, it should be celebrated. We're creating an independent integration authority with a mandate to review, evaluate, and propose changes. Transparency is not a weakness – it's the strongest tool we have.
> "Measure what works. Stop winning debates with anecdotes."
### Key Metrics
| Metric | Value |
| --- | --- |
| Annual investment | SEK 5 bn |
| Open data platform | 40+ languages |
| Responsible ministers | 15 (one per point) |
| Open APIs | 100+ |
| Monitoring frequency | Real-time |
| Total program cost/year | ~SEK 200 bn |
---
## Forward. Together.
This is a living document. The points are work in progress – better together, as always.
### Contact
[joakim@jardenberg.com](mailto:joakim@jardenberg.com)
[https://15p.jardenberg.se](https://15p.jardenberg.se)
*Licensed under CC0 (Public Domain). Free to use, share, and build upon.*