AI and GPT-3 in the age of content creation
How is AI changing the way we transform and generate content? This is the question addressed by the SlatorCon panel moderated by Florian Faes from Slator, which brought together Michel Lopez, founder and CEO of e2f, Konstantin Savenkov, CEO of Intention, and Jochen Hummel, CEO of Coreon and ESTeam.
Panelists exchanged broad views on the multi-billion dollar digital content creation market. The discussion started with translation memories (TMs) as AI learning data for a wide range of use cases, from machine translation to named entity recognition (NER).
The role of AI in transforming content – from speech to text or vice versa – was then examined, with Savenkov and Lopez arguing for a new perspective: we can see these activities as composed of successive stages of transformation, each with the potential to be improved by AI.
In closed captioning, for example, one can use Automatic Speech Recognition (ASR) to turn voice into text, change languages using machine translation (MT), and then use an AI tool to cut out. the text in chunks. In addition, one has the ability to apply AI-powered quality assurance tools at any time to improve quality.
The era of AI-powered content is upon us
The panel then moved on to the topic of content generation and the world’s largest language model, GPT-3. The speakers drew parallels between MT and GPT-3 and discussed how GPT-3 has already been put into practice.
Hummel informed the public of a fascinating short film whose script was written by GPT-3.
Below are some selected highlights from the session.
Slator: How important is a case GPT-3?
Savenkov: It’s like when Google released Neural Machine Translation in 2016 – but now for content generation. If you’re someone who feels like you’ve missed the MT wave and think you’d be doing something different next time, well, now’s your time.
Hummel: What GPT-3 does is amazing, but it also does some fun stuff. You can’t just take the content and publish it – you need a human in the loop. The parallels between the localization industry and the content creation industry are astounding.
Slator: From a practical point of view, are you already using GPT-3?
Hummel: Of course, this has already been put into practice. I recently made an angel investment in a company that automates the writing of social media posts.
Lopez: In fact, we generate more content than we translate. But it is not for human consumption. We have generated around three million words of text for the scripts, for the purpose of training the speech recognition engines.
Watch the ‘Content Creation in the Age of AI’ panel discussion and the full SlatorCon Remote event from September 2021 on demand, here.