AI for business: It’s evolution – not revolution
By Andre Torsvik, VP of Product Marketing at Vizrt Group
My newsfeed has been overwhelmed in recent months with stories surrounding the rise of Artificial Intelligence (AI) – half rapture-level technology religion and half Terminator-style doomsday prophecy, where we’ve set the wheels in motion for machines to take over the world.
Really? On 2 October 1921, the New York Times published an article called “Will Machines Devour Man”. Reviewing the book, “Social Decay and Regeneration”, the author blames all the ills of society on machines. Since then, the claim that machines are taking our jobs/destroying us/ending the world as we know it has been made so many times it does not bear counting.
A more rational way to address developments – past, present and future – would be to understand the background – the premise of the technology and what that means for the possibilities. Once done, most things are highly likely to appear as evolution, not revolution.
The difference this year is that – rather than the tech being in the background and invisible or the domain of a technological elite – suddenly lots of people can play with AI and really start to see what it is capable of. This mainstream explosion has caused some giggling in the wings as the reality is that AI, as a useable tool, has been around for years and used with great capability among businesses across the globe – including ours.
Applications of machine learning technology in the media production industry have been most successful when we have been able to leverage those things traditionally done well by machines and take them to the next level. Speed, accuracy (at the same time), the ability to ignore irrelevant data and the ability to process vast amounts of data have been key factors in solving complex problems to help our customers succeed.
Examples include numerous capabilities leveraging the strength of machine learning and AI: track fast-moving objects on the screen, from race cars to snowboarders – something that requires speed and accuracy. Instantly outline and place a whole team of basketball players and place them in virtual space to allow the camera to see things from angles it cannot actually see; something that requires attention to detail and, for human operators, quite a lot of time.
Virtually replacing the ads around a football pitch to increase ad revenues – all the while accounting for players getting in the way, fog, rain, snow, shadow and several other factors that make something like this nigh-on impossible for humans. All this is done in real-time and not post-production, and the result is a better show, higher viewer engagement and ultimately more secure revenue streams.
AI as a business tool isn’t anything new but as it develops rapidly and more companies adopt its applications, we will see changes to workplaces that can modernize business practices substantially.
So, the natural question is: Why would you deny yourself access to something perfectly suited to the task at hand? When spellchecking came along, why would anyone forego using it and risk incorrect language and typos slipping through? With AI, why would you waste time creating a structure for an argument, risk being inaccurate, and risk using less than SEO-perfect language? Why, in short, would you not pick up and use the right tool, for the right task?
And this is the crux of the matter; not until we move past the hype of AI and treat it as a tool, will we truly be able to harness its power efficiently and with precision.
Below I’ve explored just FIVE areas where AI is changing business practices and emerging as a day-to-day necessity in the media production industry for a multitude of reasons:
AI can increase business efficiencies and productivity, resulting in cost savings and accelerated outcomes for both individuals and teams across organizations, and even in everyday life. Every example listed above means higher quality at higher speeds – something that is a true unique selling point in sports production. This will reach wider; a report from the McKinsey Global Institute estimates that by 2030, AI can add about 16% higher cumulative GDP than today, or an additional 1.2 percent a year. Taking these numbers and considering them across the global economy, AI’s impact on acceleration and by extension business productivity – is substantial.
This is what is in my LinkedIn feed both generative AI and beyond. AI is increasingly being used to enhance and streamline the creative process by providing insight, inspiration, and assistance. It is a creative co-pilot always available to tech-savvy creatives and creative techies – and should not be considered a threat.
Nearly 69% of college graduates think they will lose their jobs to AI in the next years – but rather than being scared by this tool, it should be embraced and accepted as something we have been using for years. Some past, un-scary, creative AI ideas include – Spotify’s AI-powered playlist generator, which creates individualized playlists for each user based on their listening history; and Grammarly, which has been helping people write for years.
What we are seeing now is the evolution of this as a tool to enhance and speed up creativity across many job markets. We will certainly be using AI to be more creative – in the sense of creating more – and the winners in this kind of market will be the ones able to combine true creativity with savvy use of the right technological tools
In the absence of proper verification procedures, the use of AI language models, such as GPT-3 and beyond, has led to quite a few fakes and misinformation. This, combined with a widespread misunderstanding of what large language models (LLMs) really are – they are not Google search – has resulted in calls for enhanced tools and methods to verify AI-generated content.
In fact, OpenAI released a classifier to detect generated content based on the use of specific algorithms and statistical patterns – which already has been debunked as “easy to fool”. The arms race between cheaters and those trying to catch them out is not new – and will always be there. But more than that, AI is already forcing us to think about the truth in information sourcing and content origins in a profound way. And for businesses that rely on integrity in building trust, the relationship to AI must be one of proactive choices, rather than reactive misunderstandings.
The question of intellectual property rights in AI-generated content is complex and controversial, as evidenced by Harvard Business Review. Major technology companies, including Google and Microsoft, have called for standardized approaches to intellectual property (IP) rights as AI use continues to grow – especially considering the use of generative AI to generate new content based on old. This is an issue that will change how businesses work and how they interact with their customers – including the addition of specific AI usage clauses in business contracts to help deal with difficult topics around authorship and ownership.
All previous points can apply in equal measure to the writing of code and creation of software – and code is increasingly at the heart of the business value of many companies. AI can speed up the development cycle and reduce coding errors, freeing up developers to focus on more complex tasks that require an understanding of real-world application and context. It can allow quicker prototyping of products, with automated testing tools and advanced AI platforms being increasingly used to detect and fix coding errors when bringing products to market.
At the same time, human oversight is necessary to ensure the ethical ramifications of this are taken care of – what if AI-introduced code starts impacting the language of articles in your local newspaper in a certain political direction, for example, based on your assumed preferences? Do we need to label images that are not real? How is this different from a photo retouched by a human? And, at the end of the day, who do the words, images, and video it generates belong to? And whose code is it really?
While there are many profound discussions being had currently, all the way from the technical to the philosophical, this too shall pass – in the sense that AI will become more and more mainstream. And when we stop thinking of it as AI in all caps, this technology will blend into the background.
After all, it is already happening – we have been using AI for years in many of our everyday tools and mostly didn’t even realize it. Amazon and Netflix suggestion algorithms, Google has used it in search technology since 2015, it is in spam filters, automatic reply suggestions… You probably used AI many times already today without thinking “Look at me, I am using artificial intelligence!”
Once that happens, once the glamour of the ‘new’ wears off and we get into the reality of actual, business-driven applications of AI that can improve and assist our work lives daily – that is when it will get really exciting. Until then, embrace the change and use it for good. Revel in the silly 7-fingered hands generated in images, or the obnoxiously self-assured and sometimes completely incorrect factual answers you get from ChatGPT. Just remember who the tool is meant to be.