Explain the limitations of Artificial intelligence.
Background: Jam3 is a design and experience company. One of the aims of Jam3 is to generate new ideas. Those ideas are researched and developed at Jam3 Labs, thus, today’s ideas become tomorrow’s brands! One of the latest ideas is to use an algorithm to generate modifications of adidas sneakers. The type of algorithm is a “generative adversarial network (GAN)”. GANS actually make use of two different networks, one that generates new images of what is supposed to be sneakers (sport shoes), and the other tries to work out which are fake and which are real sneakers by being trained to look at real adidas sneakers. The detail of the algorithm is not important here. What is important is that the algorithm can create new trendy designs for sneakers! The visuals of the new designs of the sneakers can then be used in real campaigns and analytics can be done on the feedback (surveys, tweets, Instagram comments…..) to see which modified
Page 2 Kaplan Business School Assessment Outline 2021
designs are the most popular. From those, the popular designs, real new types of sneakers can be manufactured and sold. Sources: https://medium.com/@Jam3/how-to-make-ai-generated-sneakers-3a39464ec58d
https://towardsdatascience.com/generating-shoe-designs-with-deep-learning-5dde432a23b8
Suppose that you are in charge of a team of analysts and data scientists and you want them to adapt the algorithm (generating new sports shoes) to another specific area of fashion (e.g. dance shoes, high heal shoes, coloured shirts, summer dresses, leather coats).
Assessment Instructions
• In your own words, introduce the idea of algorithms and artificial intelligence (AI) from your workshop notes and the references below. Also explain what AI can do in a general sense. [250 words, 7 marks]
• Introduce the application area of fashion to which the team will adapt the algorithm. [100 words, 3 marks]
• Explain exactly what data you will need to gather for the algorithm, e.g. images of men’s white shorts, tap dancing shoes, race day hats, lengths, widths of items, etc. Describe what type of data this is and how that differs from other types of data, i.e. is it structured, semi- structured, unstructured? How do these types differ? [250, 7 marks]
• Describe how you will get feedback from potential customers about your new designs, e.g. tweets, and what the next steps will be once you have feedback. [200 words, 3 marks]
• Explain how you might store all of your data. [100 words, 3 marks]
• Explain the limitations of Artificial intelligence. [100 words, 3 marks]
• Include at least five references which can include one or two from your workshop notes. Use Harvard referencing. In the Academic Success Centre on MyKBS see STUDY RESOURCES – ACADEMIC SKILLS. [2 marks]
