Let’s talk about
the scary bits.

AI will change work, and some roles are at risk. “Everyone” misses the bigger picture. The World Economic Forum projects 170 million jobs created and 92 million displaced by 2030 across several global trends, a net gain of 78 million. That is a forecast, not a promise, and it is not an AI-only figure. Learning to work with the tools is a practical step.

The eerily accurate advert is more often explained by web tracking and data brokers than a secret microphone. EFF says there is no evidence of routine covert listening for targeted ads, while noting that permissions and spyware can create real microphone risks. Check app permissions and privacy settings.

Today’s chatbots generate responses from learned patterns. There is no established evidence that current systems are conscious or have human goals. Serious safety questions still matter: misuse, unreliable output and concentration of power deserve attention now.

If you can describe a task in everyday words, you can make a start. Ask for a simple meal plan or a clearer version of a letter. You can refine the answer with “shorter”, “simpler” or “try again”.

It can invent facts, citations and confident-sounding answers. Use it to explore or draft, then check claims that matter against reliable sources. Even AI providers warn users to verify important information.

Big companies have advantages, but many useful tools have free entry points. An individual can use AI to understand a form, plan a trip or prepare for an interview. Watch for subscriptions and limits before you commit.

Bias is a real risk because training data and design choices reflect the world and its inequalities. Ask who a tool was tested for, question stereotypes and keep people in charge of consequential decisions. That is a reason to use AI carefully, not blindly.

AI can make quick variations, but it does not know your taste, experience or intent. Use it for a rough start, then bring your own judgement and voice. Creators also deserve clear rules on training data and credit.

Sometimes it is. Sometimes it is convincing. There is no foolproof “AI look”, and detection tools can make mistakes. For important content, check the source and evidence rather than relying on a hunch.

AI uses electricity and water, and its demand is growing. The International Energy Agency expects data-centre electricity demand to rise substantially this decade. Its footprint depends on the tool, infrastructure and energy mix. Choose useful tasks over needless generation and ask providers for transparency.

Privacy depends on the service, its settings and what you share. Avoid putting passwords, medical records or someone else’s private information into a general chatbot. Read the provider’s controls and choose services with a clear privacy policy.

AI may help with preparation, accessibility and admin, but professional judgement and human relationships remain central. Use symptom tools to prepare questions, not diagnose yourself. UNESCO calls for AI to support educators, and medical decisions belong with qualified clinicians.