AI as the New Infrastructure According to the Stanford AI Index 2024, AI will move from the status of a “tool” to the status of “core infrastructure” in the next decade. Just like electricity in the 20th century or the Internet at the beginning of the 21st century, AI will become an integral part of almost all industries. According to a PwC report, AI can add up to $15.7 trillion to the global economy by 2035, of which:

• $6.6 trillion — due to increased productivity;

• $9.1 trillion — due to increased consumer demand driven by personalized products and services.

What will happen to work and professions?

McKinsey Global Institute predicts that by 2035, AI will automate 30% to 45% of the tasks performed by people today. However, this does not mean mass unemployment, but rather a transformation of professions:

• Routine roles will disappear: call center operators, some accountants, truck drivers.

• New professions will appear: AI ethics engineers, model trainers, digital twin designers, AI systems management specialists.

• The WEF report (2023) states: 60% of workers will need retraining by 2030 - the trend will intensify by 2035. AI will become multimodal and adaptive By 2035, most AI systems will become multimodal - they will be able to simultaneously process text, images, sound and video, as well as understand the context of interaction. Already in 2024, systems like GPT-4o appeared that “feel” the emotions of the interlocutor and adapt the communication style.

According to the MIT Technology Review, by 2035, AI will be able to:

• participate in negotiations on behalf of a person, • make autonomous decisions in business,

• create scenarios and forecasts taking into account uncertainty (not only data analysis, but also assumptions). Security and regulation The EU has already adopted the AI Act (comes into force by 2026), and similar laws will appear in dozens of countries. By 2035, it is expected that:

• all AI systems in critical sectors (medicine, finance, transport) will undergo mandatory certification,

• the presence of “black boxes” — logs of AI decisions — will become the norm,

• a global agreement on AI risks will come into force (possibly under the auspices of the UN).

AI in Science and Medicine

AI is already being used in drug discovery (e.g. DeepMind's AlphaFold), but by 2035, the Nature AI Forum predicts, it will be part of every lab. It will include:

• digital patient models — personalized medicine;

• autonomous diagnosticians — AI systems working in conjunction with doctors;

• accelerated modeling of materials, DNA, and climate scenarios.

Smart Cities and Infrastructure

According to the OECD, by 2035, the largest cities will have:

• fully automated public transport;

• adaptive traffic and energy management using AI;

• crime prediction and prevention systems (subject to ethical oversight).

Major Challenges

1. Transparency and explainability — AI must “explain” its decisions.

2. Inequality barriers – risk that countries and groups with less access to AI will lag behind.

3. Deep fakes and manipulation – by 2035, Gartner predicts that up to 70% of video content could be generated by AI, making it more difficult to verify information.

4. Threats to autonomy – the issue of control over highly autonomous AI is still open.