What are the biggest challenges of using artificial intelligence for social impact?

由人工智能和领英社区提供技术支持

Artificial intelligence (AI) is a powerful tool that can help address some of the most pressing social and environmental issues of our time. From improving health care and education to fighting climate change and poverty, AI can offer innovative solutions that can scale and adapt to different contexts and needs. However, using AI for social impact also comes with significant challenges that need to be addressed to ensure that it is ethical, inclusive, and sustainable. In this article, we will explore some of the biggest challenges of using AI for social impact and how they can be overcome.

此文章中的业界达人

由社区从 5 条内容中精选。了解更多

1 Data quality and availability

One of the key challenges of using AI for social impact is ensuring that the data used to train and evaluate AI models is of high quality and availability. Data quality refers to the accuracy, completeness, consistency, and relevance of the data, while data availability refers to the accessibility, timeliness, and security of the data. Poor data quality and availability can lead to biased, unreliable, or harmful AI outcomes that can undermine the social impact goals. For example, if the data used to train an AI system for diagnosing diseases is not representative of the population it serves, it can result in misdiagnosis or discrimination. To overcome this challenge, it is important to use data that is verified, validated, and diverse, and to ensure that the data sources and methods are transparent and accountable.

添加您的观点

James Boland

Founder and President
举报内容
Well, mate, here are a few of the biggest ones: Ethical considerations: AI systems need to be designed and deployed in an ethical manner. Access and inclusivity: AI technologies can have the potential to exacerbate existing inequalities. Trust and acceptance: Building trust in AI systems can be a challenge. Data quality and bias: AI systems rely on data to learn and make predictions. If the data used is biased or of poor quality, it can lead to biased outcomes and unfair decisions. Human-AI collaboration: Finding the right balance between human judgment and AI capabilities is another challenge. By addressing these concerns and working towards responsible and ethical AI practices, we can harness the power of AI for social good.

已翻译

赞

加载更多内容

2 Ethical and human rights implications

Another challenge of using AI for social impact is ensuring that the ethical and human rights implications of the AI applications are considered and respected. AI can have positive or negative impacts on human dignity, autonomy, privacy, equality, and justice, depending on how it is designed, deployed, and governed. For example, if an AI system for predicting crime is based on historical data that reflects existing biases and inequalities, it can reinforce or exacerbate them. To overcome this challenge, it is important to adopt ethical principles and frameworks that guide the development and use of AI for social impact, and to involve diverse stakeholders and perspectives in the decision-making process. Moreover, it is important to monitor and evaluate the impacts of AI on human rights and to ensure that there are mechanisms for redress and remedy in case of violations.

添加您的观点

Bud Hamilton
举报内容
We must be aware of the ethical and legal impacts of AI right now on organizations. For example, it was "assumed" that using AI for HR-related decisions would allow for more "objective" and "fair" decisions in areas such as hiring or performance appraisal. The problem is that AI is dependent on the data on which it is trained, and that data is often based on previous biased managerial decisions. Even when data is collected autonomously by sensors, evaluation of the output has to consider factors that might be beyond the sensor input (for example, you were late in your delivery because of unanticipated traffic). Best practice would consider an evaluation of the end-to-end data collection process, with input from all stakeholders.

已翻译

赞
Adam Rose

Senate Intern | United Nations Fellow | DEI & Social Impact Advocate
(已编辑)
举报内容
Beyond the grave ethical problem of using an inhuman machine to address inherently human issues, if the AI system is dealing with #bigdata as it would be designed for, any existing social issues not accounted and corrected for will be compounded extensively. Furthermore, while the resulting negative effects may be easy to understand, diving through the AI's decision-making process within the data to identify what information will need to be filtered or reassessed may prove more challenging given the new circumstances caused.

已翻译

赞

3 Technical and operational capacity

A third challenge of using AI for social impact is ensuring that there is adequate technical and operational capacity to develop, implement, and maintain AI solutions. Technical capacity refers to the skills, knowledge, and resources needed to design, build, and test AI systems, while operational capacity refers to the ability to deploy, manage, and update AI systems in different settings and scenarios. Lack of technical and operational capacity can limit the potential and effectiveness of AI for social impact, especially in low-resource or complex environments. For example, if an AI system for improving education is not compatible with the existing infrastructure or curriculum, it can fail to deliver the intended outcomes. To overcome this challenge, it is important to invest in building and strengthening the technical and operational capacity of the social impact sector, and to foster collaboration and partnerships among different actors and sectors.

添加您的观点

4 Social and cultural acceptance

A fourth challenge of using AI for social impact is ensuring that there is social and cultural acceptance of the AI solutions. Social and cultural acceptance refers to the level of trust, confidence, and support that the users and beneficiaries of the AI solutions have, as well as the alignment of the AI solutions with the values, norms, and expectations of the communities and societies they serve. Lack of social and cultural acceptance can hinder the adoption and impact of AI for social impact, and can also generate resistance or backlash. For example, if an AI system for enhancing financial inclusion is perceived as intrusive, unfair, or risky, it can discourage or exclude potential users. To overcome this challenge, it is important to engage and empower the users and beneficiaries of the AI solutions, and to ensure that the AI solutions are context-specific, user-centric, and culturally sensitive.

添加您的观点

Danilo T.

synchronicity. < . >
举报内容
Does Silicon Valley culture, export? Does it flow into the cultural behaviour of all the neighbourhoods of the United States? Does it flow or rather become accepted in other countries? When you observe the adoption of tech. Be it email. Be it mobile phone purchases. Be it laptop purchases. Be it Tesla purchases. In each case you are in fact uncovering people who are optimistic about the future, see value in purchasing this tech. AI will be propagate faster than anything we've observed. Moore's Law will not apply. There is no law in AI. Ai is a winner takes all mentality. Therefore whoever is prescient to pick this up, is in fact doing themselves a service. We are observing of a culture shock, as anticipated by Alvin & Heidi Toffler. Indeed.

已翻译

赞

5 Sustainability and scalability

A fifth challenge of using AI for social impact is ensuring that the AI solutions are sustainable and scalable. Sustainability refers to the ability of the AI solutions to continue to operate and deliver impact over time, while scalability refers to the ability of the AI solutions to expand and adapt to different or larger scales. Lack of sustainability and scalability can limit the long-term and widespread impact of AI for social impact, and can also create dependency or vulnerability. For example, if an AI system for reducing food waste is dependent on external funding or support, it can cease to function or lose its relevance if the funding or support ends. To overcome this challenge, it is important to design and implement AI solutions that are cost-effective, resilient, and adaptable, and to leverage existing networks and platforms that can facilitate their dissemination and replication.

添加您的观点

6 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?

添加您的观点

Social Impact

+ 关注

给文章评分

我们借助人工智能创建了此文章。您认为这篇文章怎么样？

很棒不太好

举报此文章

What are the biggest challenges of using artificial intelligence for social impact?

1

2

3

4

5

6

1 Data quality and availability

2 Ethical and human rights implications

3 Technical and operational capacity

4 Social and cultural acceptance

5 Sustainability and scalability

6 Here’s what else to consider

Social Impact

给文章评分

感谢您的反馈

更多Social Impact相关文章

更多相关阅读内容