What is artificial intelligence?

Artificial intelligence or AI refers to machine-based systems that can make predictions, recommendations or decisions based on human inputs. The predictions, recommendations and decisions made by AI can affect real or virtual environments.

Artificial intelligence encompasses many different fields and approaches and covers a wide range of applications. Important concepts related to AI include machine learning, deep learning, natural language processing, computational problems, heuristics and complex algorithms.

In short, artificial intelligence aims to create machines that can:

1. Learn: The ability to improve performance through experience.
2. Deduce: The ability to make reasonable judgements and decisions based on the available information.
3. Adapt: The ability to respond and adjust to changing circumstances.

Artificial intelligence has a wide range of applications, including voice and image recognition, natural language processing, robotics, game design, healthcare, transportation, security, business and many other fields.

It is important to note that artificial intelligence is not a single technology or method, but rather an umbrella term that covers many different approaches and techniques. Applications of AI range from simple tasks such as face recognition to more complex tasks such as guiding self-driving cars.

In the past, there was a big difference between search engines (e.g. Google) and AI language models. Nowadays, many AI programs also include a search function.

• As a rule, search engines primarily look for websites.
• AI applications are typically based on language models. They search for information in large data sets, combine different pieces of information and present their results in a ready-to-use format.
• Perplexity AI is a combination of ChatGPT and a search engine that searches for answers to user questions in real time across the web.
• The paid version of ChatGPT combines AI with search capabilities. It retrieves information and provides sources while generating text with its language model.

1. Speech recognition software: These programs use artificial intelligence techniques to recognise and interpret spoken language. They can, for example, convert speech to text or recognise a speaker's voice cues. 
   • Elevenlabs – an AI-based text-to-speech tool that is available for many languages.
2. Machine translation applications: AI-based translation applications can automatically translate text from one language to another without human intervention. These applications are often used to translate web pages or handle international affairs. 
   • Grammarly.com – an AI-based tool that can be used to improve texts.
3. Image recognition software: These programs use artificial intelligence to recognise and analyse images. For example, they can identify objects, people or animals from images. Image recognition software is used in applications such as security cameras, medicine and automated vehicle detection.
4. Recommendation systems: These programs use AI technology to analyse information about different products or services based on user preferences. They then generate personalised recommendations to help users find the right products, movies or music.
5. Chatbots: AI chatbots (e.g. OpiqGPT, ChatGPT, Copilot, Google Gemini) are programs that can communicate with humans in natural language. They can answer questions, give advice and provide customer service in a variety of fields, such as e-commerce or banking.
6. Image creation and editing: Dall-e, Canva Pro, Playground
7. Content creation: Scalenut, copy.ai and Scribe are examples of AI-based content creation tools.
8. Video creation software: Tools such as Invideo AI can be used to make incredible videos in just a couple of minutes. Also see this demonstration video of OpenAI's Sora tool.
9. Many kinds of professional software: 
   • Code writing and editing tools 
   • Marketing tools 
   • Game design tools 
   • Forecasting tools (used e.g. to predict stock market prices) 
   • Tools and applications used to help in health care

Sources: Faktabaari and ChatGPT

Everyone should understand what artificial intelligence is, how to approach it and how to use it. This is called AI literacy.

1. Not all AI applications comply with EU data protection requirements.
   ​→ Careless use can lead to data leaks to third parties.
2. [annotation: Ohjelmien eikä ihmisten manuaalisesti käyttämät käyttäjätunnukset.]Bots[annotation: Users controlled by software rather than people.] control social media in many different ways. For example, they influence social networking algorithms by generating large numbers of views. → As a result, the algorithm mistakenly considers the targeted posts to be popular. → The social networking service shows more and more of them. → This leads to millions of people around the world seeing social media posts whose visibility is deliberately controlled by bots. → Thus, our opinions can be swayed quickly and on a global scale.
3. The material used by AIs is rarely carefully selected. Because of this, you cannot rely on information produced by AI. The AI cannot judge whether the information it provides is true or not.
4. AI can produce many kinds of negative things, such as fake videos (deepfakes) whose origin is almost impossible to know.
5. It is important to understand that the use of AI will influence people's opinions and skills. Threats include the possibility of false information becoming public opinion, hybrid threats, automatic censorship and the acceptance of false information as truth. It is therefore increasingly important for students to learn how to think critically.

1. Prohibited uses: e.g. monitoring pupils with facial recognition software in surveillance cameras.
2. High-risk uses: e.g. student selection, test scoring and the use of people's personal data. It is important to understand that feeding data into an AI program is equivalent to making the data public. Decision-making should not be delegated to AI, but AI can be used to support decision-making, for example through analysis and visualisation.
3. Low-risk, high-mass applications: e.g. customer service bots.
4. The teacher can define how students can use AI to complete various tasks:
   1. The use of AI is forbidden. Think in advance about how the ban is enforced and what the consequences of getting caught will be.
   2. The use of AI is permitted for information gathering and other equivalent tasks. AI-based applications are not normally cited as references, as they are brainstorming and writing tools, not factual sources of information. However, if it is necessary to reference AI-generated text because it is relevant to the work, the correct reference template is as follows
      ​​OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat
   3. The use of AI is allowed in all use cases.

• Responsibility: The author is always responsible for his/her own work.
• Knowledge: AI systems are just programs and they have their limitations. AI has no knowledge or understanding of the content it produces. Be aware of these limitations so that you can assess the suitability of using AI in different situations.
• Is the accuracy of the result important? If the answer is no, then you can use AI with confidence. If the answer is yes, consider whether you have the expertise to verify the information provided by the AI. If you do not have the necessary expertise, you should not use AI in the given context.
• Ethical principles: Follow general ethical principles such as fairness, equality and respect for other students and teachers. (Sources: Arene and Tavastia)
• Do not break copyright laws!
  1. Make sure you have the right to use the material as a prompt. All input will be used by the AI program and may therefore be used by third parties.
  2. No court cases have yet been decided on the issue of copyright in AI-generated images or other content. Because of this, you should be careful when using AI tools to create content.