Future of work

Is X a good career after AI?

The honest answer is almost never yes or no. AI removes tasks, not job titles — so the useful question is which parts of your work just got cheap, and which parts got scarcer. Below is a role-by-role read: what is being automated, what stays durable, and the move that keeps you valuable. Written from Durban, so the South African picture is included rather than assumed away.

Find your role

Search a career or filter by how hard AI is hitting it

Software developer

Reshaped

Still a good career, but the entry rung is higher. Writing routine code is now the cheapest part of the job; judgement about systems, data and correctness is the expensive part.

What AI absorbs: Boilerplate, CRUD screens, test scaffolds, first-draft functions, routine debugging.
What stays yours: System design, security, performance, integration with messy real-world data, owning outcomes in production.
The move: Ship things end to end. Depth in one domain (payments, health data, energy) beats generic full-stack breadth.

Lawyer / legal professional

Augmented

Good career, with the leverage shifting to those who can supervise AI output. Drafting speed is no longer a differentiator; accountability for advice is.

What AI absorbs: First-draft contracts, document review, summarising bundles, routine research memos.
What stays yours: Advocacy, negotiation, regulatory strategy, taking professional responsibility for the answer.
The move: Learn to verify and govern AI output. Legal-tech, IP and AI-regulation work is where new demand is forming.

Data analyst

Reshaped

Mixed. Dashboard-building and query-writing roles are compressing; roles that define what should be measured are not.

What AI absorbs: SQL generation, chart building, routine reporting, basic cleaning.
What stays yours: Framing the question, judging data quality, causal reasoning, telling a decision-maker what to do.
The move: Move toward decision science, data engineering, or a specific industry where the data is proprietary and hard to get.

Accountant / auditor

Augmented

Solid, because the work is regulated and someone must sign. Automation eats the preparation, not the attestation.

What AI absorbs: Reconciliations, classification, first-pass variance analysis, drafting notes.
What stays yours: Assurance, professional scepticism, regulatory sign-off, advising on structure and risk.
The move: Add assurance over non-financial data — ESG, carbon and supply-chain disclosure is a growing mandate.

Graphic designer

Reshaped

The hardest hit of this list for commodity work. Generative tools produce acceptable assets instantly; brand systems still need a designer.

What AI absorbs: Stock imagery, social variants, simple layouts, iteration on a defined style.
What stays yours: Brand strategy, art direction, taste, working with clients who cannot describe what they want.
The move: Sell direction and systems, not assets per unit. Product and motion design hold value better than static output.

Teacher / lecturer

Resilient

Resilient. AI changes assessment and preparation far more than it changes the job of getting a room of people to understand something.

What AI absorbs: Lesson-plan drafts, marking rubrics, worked examples, differentiated materials.
What stays yours: Motivation, classroom management, pastoral judgement, assessing genuine understanding.
The move: Own AI-era assessment design — the schools and universities that solve this will hire the people who already have.

Nurse / allied health

Resilient

Among the most durable careers on this list. Demand is demographic, and the work is physical, licensed and relational.

What AI absorbs: Documentation, triage prompts, scheduling, first-pass imaging flags.
What stays yours: Hands-on care, clinical judgement under uncertainty, patient trust, working in under-resourced settings.
The move: Specialise, and take the documentation-automation skills with you — that is where the time savings land.

Customer support

Reshaped

The weakest case for a long career at tier one. Volume handling is being absorbed; escalation and trust work is not.

What AI absorbs: FAQ answering, ticket routing, order status, first-response drafting.
What stays yours: Angry-customer recovery, complex claims, regulated complaints, designing the support system itself.
The move: Move to support operations, quality, or the team that builds and audits the automated assistant.

How to read the labels

Reshaped: Large parts of the current job are being automated; the role survives in a changed form.
Augmented: AI speeds the work up, but accountability and licensing keep the role in human hands.
Resilient: Physical, relational or regulated work that automation reaches slowly.

Questions people actually ask

Will AI take my job entirely, or part of it?

For almost every role on this list, tasks are being automated rather than whole jobs. The risk is not disappearance — it is that the tasks you are paid most for become cheap, and your bargaining position falls with them.

Is it too late to start a career in tech?

No, but the old entry path — get hired to do simple work, learn on the job — is narrowing, because simple work is what tools do best. Enter with a portfolio of finished, working things rather than certificates.

Which skills hold value?

Anything where being wrong is expensive and someone must be accountable: safety, regulation, money, health, infrastructure. Also anything requiring physical presence or trust built over time.

Does this look different in South Africa?

Yes. Local demand is shaped by an economy with high unemployment, constrained electricity and logistics, and a strong regulated-professional sector. Roles tied to compliance, energy, ports and public infrastructure are less exposed to remote automation than globally traded knowledge work.

What should I do this quarter?

Pick one task in your job that AI now does well, and get good at supervising it. Then pick one task it cannot do, and make that the thing you are known for.

Where to check this yourself

Primary sources, so you can verify rather than trust

Signal tracks these shifts daily — the future-of-work thread runs through today’s briefing and the South Africa view. You can also ask Signal about your own role.