AI See You

07/12/2024

You may have heard the word 'deepfake' thrown around recently. If you're unfamiliar with the term, they are audio, images, or videos of people who have been digitally altered using AI to act or speak in a way that differs from reality. 🔊🖼️🎥

You may have used some of these tools yourself:

Ever colourised an old photo with an automatic online software? 🖼️

How about using a text-to-speech tool to read aloud an article you're looking at? 🗣️

While these tools are impressive, we cannot ignore the dangers. Deepfakes are often used with malicious intent, and they are increasingly convincing. ⚠️

Have you seen any concerning deepfakes online recently? How worries are you about their risks to democracy in the hands of bad actors? Take a look at some of the dangers below and share your thoughts!

24/11/2024

How closely do the experiences of the students we interviewed recently align with broader trends? ...Pretty closely it seems...

AI, when used responsibly, could lift the burden on markers, create new ways to teach and learn, while freeing up teachers' time to deliver better and more engaging study programmes.

However, with these opportunities, come some major hurdles:

Reliance on AI could hinder the development of critical thinking skills and reduce student motivation; the instant nature of AI responses are likely to generate only surface level understandings of core topics, reducing scope for analytical skills, information literacy and original thinking. Students may also feel less inclined to engage deeply with their research, limiting their potential to 'go beyond'.

More time interacting with AI, and less time in the classroom, could also stunt the development of soft skills such as teamwork and debate; lower levels of student engagement post-Covid are being compounded by reliance on AI technology, reducing opportunities to practice communication, leadership, and active listening skills.

AI also threatens to damage academic integrity and fairness across the board; research from the UK's higher education think-tank HEPI reveals that students from disadvantaged backgrounds are less likely to access AI tools than those in the most privileged backgrounds. The lack of clarity on AI misuse between institutions threatens to further contribute to this divide.

24/11/2024

Controversies regarding AI keep popping up in the news, but what do they have in common? 💭

For all its impressive abilities, AI is only as good as the data it is trained on. If that data is biased, incomplete, or inaccurate, the output will reflect that. 📊

How confident would you be if you knew AI was involved in recruiting for your dream job? What about if AI were to serve on a jury? Would you let AI make decisions of life or death? - That's already happening with driverless cars... 🏎️

While the negative consequences of these situations seem far fetched, we really just don't know how serious they could be...

This is largely due to the deficit of transparency around how AI companies gather data, where they draw it from, and to what extent it is vetted and checked. 🪟

Ultimately, data used to train AI should be:
- Diverse,
- Representative,
- Free from unchecked biases,
- Obtained with the knowledge and consent of the original creators where applicable.

21/11/2024

How do students use AI in their studies? How do they think it impacts their learning? We asked some undergraduates to share their experiences...

21/11/2024

Last week, we got you to start thinking about AI generally, its impact on you, and the misunderstandings that commonly play out in public discourse. 🤖

This week, we're honing in on several key at-risk areas, and how AI could threaten to change them for worse. ⚠️

We'll begin with education: Through an interview with two undergraduate students, we explore how AI tools are being used in education, and how this may threaten academic integrity and curtail the development of critical thinking and research skills. 👨‍🎓

We'll also consider the practical and ethical issues of AI data usage, what this means for your privacy, and whether we can trust the information generated by AI assistants. 📈

In a look at the more nefarious, we'll consider how AI technologies such as deepfakes and voice replication tools are increasingly used as tools to spread disinformation, threatening to shake the bedrock of our democracy. 👼👺

Finally, we'll consider how the exponential development of AI technology could threaten to detail entire careers and industries, as automation becomes an attractive alternative to paid employees. - We'll be taking a specific look at the creative industries, with insights from those working in the field. 🎸🎨

17/11/2024

17/11/2024

Are you fully aware of how AI impacts your daily life? Recent studies from the UK's ONS suggest otherwise... 😳

68% of those over 70 reported not using AI, yet as the evidence shows, we might be less knowledgeable of where AI actually operates than we imagine. ⤵️

Wear a smartwatch or fitness tracker? These use biometric data to track your exercise and sleeping patterns. 🧘😴

What about home security systems? These increasingly use facial recognition technology to spot intruders. 🔐📸

How about streaming? Watch Netflix or listen to Spotify? These services suggest new content for based on your habits. 🎧📺

No? Hmm. Do you shop online? Sites like Amazon and eBay use data from your prior purchases to recommend products. 🛒💳

Alright, one more: email spam filters. These employ AI technology to analyse message content and detect harmful / unwanted mail. 📧💻

How aware are you about AI and how your data is used? 🤔

17/11/2024

According to a 2022 PewResearch survey, we might be less aware of how AI impacts us than we think... 🤔

Less than a third of Americans considered themselves regular users of AI... ...44% thought they did not interact with AI at all.

There's clearly a public awareness deficit around the uses of AI.

Do you consider yourself a user of AI? Do you see anything you interact with below? Is there anything that surprises you?

Let us know! 👇