Careers and the outlook to 2030
Lesson 23 of 23 in the course “Quantum”. Free, with animations and a quiz.
Jobs in quantum
The quantum industry isn’t just physicists in labs. There are four big directions. Quantum software and algorithms: you need Python, linear algebra and libraries like Qiskit. Post-quantum migration: this is regular cybersecurity, no physics needed, and demand is growing thanks to the 2030–2035 deadlines. Hardware: cryogenics, microwave electronics, lasers and optics, usually with an engineering or physics degree. And product, analysis and content: people who understand the tech and can explain it are welcome everywhere. After this course you already have a base for that last one.
Where to start for free
Everything you need to start is free. IBM’s Qiskit is the most popular library, and on the free Open Plan IBM gives you 10 minutes on real quantum computers every 28 days. That’s enough for dozens of small experiments. Google’s Cirq is handy for research, Xanadu’s PennyLane for quantum machine learning, and Nvidia’s CUDA-Q for hybrid GPU-plus-quantum computing. They all run on Python. And don’t forget this course’s trainers: you can play with circuits and the Bloch sphere right in your browser.
A 90-day plan
Want to try it seriously? Here’s a realistic plan at 30–40 minutes a day. Month one: Python and basic linear algebra, vectors and matrices, exactly what describes qubits and gates. Month two: Qiskit. The goal is to build a circuit with H and CNOT and run a Bell state on a real IBM computer. You’ll see how noise spoils the perfect 50/50. Month three: your own project. For example, Grover on 3 qubits, or a script that checks whether websites support post-quantum key exchange. Put the code on GitHub and write a post about what you learned. That’s a portfolio you can take to an internship.
What’s promised for the 2030s
Let’s collect the key dates. IBM promises the Starling machine by 2029: 200 logical qubits and 100 million gates in a row. Quantinuum plans Apollo, a fully fault-tolerant computer, for the same year. In 2030, under NIST’s draft, RSA and elliptic curves become deprecated. And DARPA’s QBI programme wants to check by 2033 whether any company can build a machine whose value exceeds its cost. Remember the rule from level 4: these are plans. Check them against what companies have already delivered. IBM, for example, has broadly hit its intermediate steps so far.
What experts think
So when is Q-Day? Honest answer: nobody knows the exact date. The best we have is the Global Risk Institute’s regular survey. In its 2025 report, 26 experts put the chance of a quantum computer able to break today’s cryptography at 28–49% within 10 years and 51–70% within 15 years. That’s not “never” and not “tomorrow.” It’s like a weather forecast with a 40% chance of rain: worth taking an umbrella. For crypto and data, the umbrella is moving to new addresses and PQC ahead of time.
Your quantum wrap-up
You’ve gone from “what is a quantum” to the race map and the quantum risks to your money. Now you understand why a qubit isn’t “0 and 1 at once” in the everyday sense, why a quantum computer doesn’t try everything at once, what error correction is and why qubit count alone means nothing. Take three rules with you. Check headlines with the checklist. Guard your keys: fresh addresses and updated wallets. And don’t buy hype: no “quantum-safe” tokens, no “Quantum AI” bots. And if you want more, you have a 90-day plan.
Check yourself
- Can you get into the quantum industry without a physics PhD?
Correct answer: Yes, for example via PQC migration, quantum software or product. PQC is cybersecurity, software is Python and linear algebra, and product needs people who understand the tech. - How should you read a promise of “200 logical qubits by 2029”?
Correct answer: As a plan worth checking against steps already delivered. A roadmap is a claim on the future. Its reliability shows in how the company delivered past steps. - Experts give 28–49% odds of a code-breaker within 10 years. What’s the right takeaway?
Correct answer: The risk is real: prepare now, but without panic. Nearly a coin flip within 10 years is a serious risk for data that lives longer. Hence the 2030–2035 deadlines.