About 4 months ago, I started managing product design organizations for Facebook's artificial intelligence team. We are a central organization that provides AI services for Facebook, Instagram and other Facebook applications. We are also committed to developing new experiences driven by AI.

Before joining the AI ​​team, I have been reading a lot of knowledge about artificial intelligence and I am very excited about working in emerging areas. In the past 10 years, I have been researching mature mobile and web applications, so I feel really energetic in a learning and blurry position. I also feel responsible for knowing that artificial intelligence technology will revolutionize society and even surpass the mobile revolution. I want to use my skills to make sure that AI tools are built in a responsible way.

To be honest, I am not entirely sure about the role of product design in the AI ​​team. We don't have our own consumer product surface, so the team doesn't have a daily maintenance app. I also have a preconceived bias that opposes designing “technical priorities” and “human problems first,” so figure out how to work effectively in a world where technology capabilities are real-time (and you are trying to apply them in the right way) I feel depressed and unfamiliar.

I think it might be helpful to share with others the methods we use to provide design support and create new experiences for developers and application team partners. This information may be useful if you are building an AI design team, or if you are curious about the role of designing an AI.

In general, designers on the team are working on the following project types:

Design AI prototype

When AI technology teams need prototypes, demos or visualizations of their technology, they work with product designers. AI designers ensure that people see the possibilities of AI. We create prototypes that show how people use specific technical capabilities when AI is working well. As an example, we might create an AI demo to provide a possible title for your Instagram post, or AI to help you know where to buy a pair of shoes you see in your post.

For AI demo designers, the creation process typically describes a set of product ideas as a one-page newsletter with a set of visual effects that outline the idea. We use a pager to gain recognition from a wider team, and this idea deserves further expansion. The completed prototype is used to green a new research project or development project within the AI ​​organization. 
 
Designers working on AI presentations are working on a variety of ideas, and they are very productive, not deep into a technology field. They have a deep understanding of current AI capabilities and track how new features are developed. As a designer working on a presentation, you may be creating a visual search experience where visually impaired users can “view” news feed posts by scrolling through their photos.

Shaping new technology

AI designers can also serve as the embedded part of the technical team, shaping new technologies. AI designers are committed to long-term investment areas such as computer vision, voice, language, video and AI assistant programs. These project teams are developing new AI functions (rather than using existing AI functions). Your work as a designer on the technical team blurs the boundaries between design, PM and research. You are working to ensure that the developer team understands the user needs associated with a particular type of technology and is building technology to directly address these needs. You are imagining the user experience that users may see in 3 – 5 years, and you will make significant progress in artificial intelligence technology. AI designers have a deep understanding of the long-term vision of artificial intelligence in their areas of concern.

Develop AI-centric products

Another project that AI designers may be working on is designing new AI-centric applications, testing their viability, and (if hopeful) using them for further investment. Projects in this area have a much longer timeframe, so you work in a world where you imagine people living 5 to 10 years later. You are thinking about how artificial intelligence can provide better social services for people in the long run. The product you build may be a future Facebook or Instagram, or it may be a new way to interact with your community.

Designers in this field often have a background in building new hardware or using new application ideas in innovative spaces. They are satisfied with the blurry sex and understand how to create standards that continue to pursue or close the line of exploration.

Collect data for AI learning

One of the most important parts of shaping AI technology is to provide the right data for AI to learn. AI designers work with engineers who work on this data collection and annotation build tool and design a platform that supports the tool to simplify efficiency and make it intuitive to collect high-quality data. In some cases, designers help collect data sets when more automated methods don't work.

Here's an example of how designers can help with data collection to help with AI learning: We may need to teach AI to know when a person raises their hand. To teach artificial intelligence, we collected thousands of video examples of people raising their hands to ensure that examples come from people of different ages, genders, races, and physical differences so that the data set is inclusive. Then, they gathered thousands of people to help the AI ​​know when to answer "this person raised his hand" should be "no." Finally, they showed examples of raising their hands. It may be harmful (that is, someone raises a Nazi tribute) so that artificial intelligence can learn to mark these examples when identifying.

AI designers help ensure that we develop data collection systems that are safe, fair, ethical, and applicable to real-world user issues. These designers create frameworks and guidelines as needed to ensure safe and reliable use of AI. They designed the libraries that our machine learning system uses to learn to make sure they are unbiased, clean and effective.

Design AI developer tools

Designers of the AI ​​team can also build applications for AI engineers and the external developer community through open source projects. The team's product designers are building specialized developer applications to ensure that we create a set of tools that work well as systems. Designers in this field need to have an in-depth understanding of the artificial intelligence development process and are passionate about supporting engineers around the world. They usually have a strong background in engineering or system thinking.


At a very high level, AI designers are bridging the gap between user needs and technical capabilities. We know that our goal is to empower people to build communities and bring the world closer, and we want to harness the current and future capabilities of artificial intelligence to design the future of how people connect and collaborate.

The roles of a typical product designer and an AI designer are quite different. Our output usually clearly shows how the engineer applied the technology, rather than launching the product features shown in the application in an immediate and obvious way. Since AI functions may take 2-3 years to develop, it is very important for designers to help developers understand the potential of different solutions and their impact on people's lives when developing AI. What's more exciting is that in addition to using existing AI functions to solve the problems that exist on today's platforms, these designers can also imagine and imagine how AI will reshape the world of the future.

This article is transferred from medium,Original address