How can AI and Machine Learning (ML) be used to help streamline the back office or network diagnostics, crunch big data and or aid better decision making?

Chris Hodges: We see artificial intelligence, machine learning, and robotic process automation as an integrated suite of solutions, which has proven successful at dramatic improvements in back in the office and operational level processes. Specifically, robotic process automation is directly applicable to routine rules-based back and operational processes, which currently consume enormous amounts of human resources and time. Automation of routine and rules based tasks like report generation, account entry, account closure, media uploads and a wide variety of financial processes has three immediate and direct benefits. Firstly, the automation makes the process go faster. Secondly, the automation reduces the errors ubiquitous in swivel-chair processes where a person goes back-and-forth between multiple systems with mind numbing repetition. Thirdly, automating these back office processes frees up human capital to do more creative and human level tasks.

Some specific examples in the media industry are media uploads, reformatting, renaming, re-organization and responding to network alarms interruptions and outages. All of these are well suited for applying intelligent automation.

Once processes are automated they produce an enormous amount of real-time data, which can be analyzed for patterns and trends and allow for preemptive process changes or adjustments.

How can AI/ML impact the creative or editorial process of a production today?

CH: Artificial intelligence can analyze successful video or media projects based on specified criteria. These criteria might include the number of people, shot angle, gender, movement, sound levels, etc. All of these can be analyzed by artificial intelligence and machine learning producing patterns of production for specific success criteria and markets. Today, this is done largely through a combination of producers with particular individual personal styles. AI/ML will allow this to become more systematic, repeatable and scalable across multiple projects all at once.

What are the current limits of AI/ML in media?

CH: Today, artificial intelligence and machine learning combined with robotic process automation will certainly allow processes is to be done quickly and to help produce projects which conform and fit to specific targets. What artificial intelligence and machine learning are unlikely to solve anytime soon our challenges which require empathy, social intelligence and creative intelligence.

Over time – even this intelligence will be quantified and built into future processes. For example rules learned for a chat bot might look like this: “When suggesting a new path or option on the phone to a client ‘perhaps’ elicits a more favourable mood and attitude – so use it, but only so many times per call and with only certain cultures, etc.”

What should media companies be doing to incorporate AI/ML into their operations?

CH: The successful implementation of AI/ML into media organizations is similar in many ways to implementation in any other organization. However, there are many differences in that creative people tend to believe everything they do can only be done by them.

The most successful implementations follow an interactive process. The first step is to understand the potential of a variety of tools by seeing previous semi-related or closely related examples. This opens the door to the creative process and connects the subject matter experts within the media company to the potential of the technology.

Next, companies make a deliberate attempt to ‘discover’ potential opportunities where machine learning artificial intelligence is the appropriate solution. The next step is generally a proof of concept, which is less about testing the technology and more about the change in readiness process for an organization. Media types need to see how the machine will help them do their job better and not be a threat to what they value most, i.e. the creative process.

Most companies simply don't have the internal resources or experience to be able to take advantage of AI/ML yet. The trade-off becomes build or buy and if time is of the essence relying on external resources for much of this work is the most common path to success.

What will AI be able to do in 2022 that it can’t do now?

In the future, the dream would be for entire production plans and creative plans to be fed by artificial intelligence and machine learning patterns which have been proven to work in specific customer examples. The benefits will be that the work is done faster, the work will better hit the desired target market, elicits the desired response and is done at a lower cost and with more consistent quality.

Any pattern that can be found can be replicated. The biggest drawbacks today are the lack of existing data to feed the AI engine and lack of understanding how the AI engine can help. Companies need to start crawling now to walk next month and to run next quarter. Each success is built on previous successes and the expansion of internal thinking. Once you see that a spreadsheet can add numbers it doesn’t take long until you ask it to handle budgets, business plans and semi-automated reports.