AI For Social Good

The recent expansion of Artificial Intelligence AI or as I like to call it the AI revolution directly impacts multiple aspects of human life, this is primarily due to the importance of AI in solving complex tasks that would regularly require humans additional time to undergo.

Unlike how most people view it this is not bad for the economy or the resilience of companies worldwide, on the contrary as a computer science student I view that this would be a great opportunity to mitigate a new generation of work perspective, where one has to learn how to collaborate with technology to be able to keep up with today's development in the world. For example, this could translate into instead of hiring a cleaner you would have to hire a person that could program the cleaning AI-managed bots at the company.

Companies such as 英伟达 and 台积公司 continue their rapid technological advancements in the chip manufacturing industry adding more powerful computational power to GPUs and CPUs. This mainly contributes towards enabling more complex and advanced AI models with high-accuracy prediction and advanced neural networks.

Supervision & Maintenance

Any type of modern AI in the world would require constant supervision even if the AI/Machine is running on an Application Programming Interface API model rather than an AI model of its own, consistent supervision and monitoring are required to be able to ensure the effectiveness and efficiency towards the model achieving the required or desired cooperation or personal goals.

Nonetheless, maintenance is also a critical aspect of maintaining the effectiveness and operation of artificial intelligence in general any model would require constant updates to ensure bug fixes, error mitigation and even more. The more an AI works the more it learns think of it as a child who is continuously learning the more you teach it the more it develops.

Since AI has just recently begun its uprising with some models carrying high risks but not as high as to reach the critical point in consequence scale. Some AI models almost have zero risk scale, since it mainly does not work towards prediction more than analysis and finding accurate answers or locating specific data in a data set such as Google Cloud Vertex AI.

Social Uses of AI

Healthcare:

AI can be used in the applications of healthcare algorithms and analysis of data, from the detection of specific cells in the blood, analyzing mental health patterns of stress anxiety and depression and improvement of patient outcomes, treatment plans, prediction of specific deterioration patterns of health and particularly benefiting medical research.

Companies such as DK AI Lab are starting to utilize the computing power provided by major companies worldwide and apply scientific and medical research to the development of AI in a variety of healthcare implementations. I have had the privilege of interviewing the CEO of the company Tarry Singh who has highlighted when discussing the topic of social good of AI and healthcare "Don’t underestimate the power of technology". He even highlighted the great potential of the continuous technological advancements in the field and predicts that in the near future, humans would be able to interact with augmented reality operated AI models.

Education:

AI is reshaping the concept of education worldwide, there is no need for marketing for AI to reach every single student in university in 2024. Practically, almost all university students are knowledgeable of the existence of AI models such as the ChatGPT OpenAI model and Gemini AI 谷歌 model. This does not need reference whereas each university has updated its policies worldwide when it comes to plagiarizing using an AI model. The ironic part is that it's not even getting easier for them to detect plagiarism from such models because they themselves use AI also to detect plagiarism anomalies in text which in the near future can be potentially exposed.

However, let's not focus on one side of the coin, whenever we as humans need to learn we need to read from a book or let tutors mark our exams or quizzes and usually wait for feedback on the answers provided which can give us an explained idea of the wrong answers and mistakes made. Continuously, we turn to reading and solving more problems to gain more knowledge requiring the tutor to explain once again to finally understand the depth of needed knowledge and reach the desired answer.

Fundamentally, this is why AI models have quickly gained their current fame, due to their moderate to somewhat adequate accuracy when it comes to the body of knowledge in a variety of topics. Indeed the dependence on such models cannot be entirely be taken for granted, up till today there are still bugs and issues that have been discovered and some of the prompts generated from the highest moderated AI models. Moreover, there has been already plenty of research by academics and scientists in the making specifically towards the accuracy and prediction of correct outcomes when using AI models.

There are companies out there such as Reality AI Lab , which are currently developing multiple educational AI models using APIs such as Google Gemini that can be utilized by primary school and high school teachers, aiming for the future to potentially university professors.

The future of AI in the educational fields is bright and looking very promising even right now professional programmers use the assistance of AI models to be able to develop their programming and coding skills using AI as a backbone for their memory and recall of complex data sets, math fundamentals or syntax.

Management:

There is no denying that AI has magical capabilities of organizing large sets of data in really short amounts of time which makes it an ideal tool to be utilized in organizational management. Management requires social multilingual understanding and cross-cultural communication with AI powers translation tools it breaks communication barriers across the globe, creating a real-time translation of multiple spoken and written languages.

The use of AI can proactively contribute to the automation of mundane tasks providing more concentration towards strategic objectives. This allows a smoother process towards tasks division scheduling and management, boosting team productivity and providing effective and efficient allocation of company resources increasing overall resilience.

The impact of AI management particularly participates in data analysis leveraging its capabilities to make informed decisions with high accuracy towards the discovery of anomalies using AI algorithms that detect patterns and colorations that may be somewhat challenging to detect by human analysts. AI is not new to management within organizations it's used in almost all programs and applications such as Microsoft Excel which uses AI to organize most fits equations and complex drawings.

Environmental Sustainability & Safety:

it's popular nowadays that smart cities would be optimizing the usage of resources towards the reduction of waste. Smart grids using the Internet of Things technologies and implementations could be powered by an AI manager to distribute energy effectively, reducing carbon footprints and crop yields as well as provide analytics in regards to soil health and water optimization that can be utilized by farmers in their practices.

Moreover, AI would help in the detection of high risks not just in the environment but in private and governmental companies allowing safety officers or inspectors to be able to use AI to assist in the analysis and control measures set for each risk detected this would provide higher effectiveness and efficiency in ensuring occupational health and safety measures.

Imagine a city where traffic lights adjust in real-time to the flow of traffic, waste management systems signal when bins are full, and streetlights dim or brighten based on the presence of pedestrians.

Conclusion:

The need for AI is becoming more and more critical to our daily use. It is a breakthrough in education and the way we handle and perceive technology. However, the need for supervision is essential for detecting anomalies bugs and errors are being discovered and fixed almost every day behind the scenes of building each AI model.

Utilizing AI models is different from programming and building from the ground-up AI models. One could create their own private AI that could be used for specific needs for detection of certain outcomes. Nonetheless, programmers could use already built AI models and mitigate them into newer ones, as a programmer myself I could describe this as it's like being able to borrow your friend's Wi-Fi he gives you Wi-Fi but you get the ability to browse all over the Internet however you like.

To conclude, AI efficiency mainly depends on how well the AI's neural network can be constructed, the better its layout and shape the better its required outcome precision. Being able to use AI, however, is more critical than being able to build it. Thus, models that have recently shown up require the usage of experts.

This is mainly because AI initially cannot conduct tasks on its own without the supervision or the prompt of an actual expert or professional within the operated field. For instance, a doctor would have to be well knowledgeable of medical science to be able to use an AI that targets cancer cells in the bloodstream, if an average Joe wants to use this AI he would be able understand to the outcome of the requested query. Resulting in not detecting if there was an analytical mistake in the given or outputted data or if the AI is giving the needed accurate results.

Resources

• Vertex AI Model Explanation
• Google Gemini AI Model Info
• OpenAI Model Info
• AI & Future of Humans Report
• Human Compatible: Artificial Intelligence and the Problem of Control, 2019 by computer scientist Stuart J. Russell.