Reclaim TikTok: An inspirational AI solution to fight fake information

Today we speak with Matthias Killer, one of the Team Leaders (M.Sc. in Robotics & AI with prior founding and team lead experience) of Reclaim Tiktok, the winning team of our last Open Innovation Programme who pitched at re:publica in May this year.

N3XTCODER: Reclaim Tiktok is the name of your team, why did you choose it?

Matthias: We felt inspired by the official initiative/hashtag on TikTok, so we use it for the programme to be really clear on what our objective was.

Can you tell us a bit more about your project, so the readers can know about it?

Sure! The official goal we were targeting was levelling the playing field of political communication on TikTok for democratic organisations by using AI-enhanced Analytics and by gaining insights into the success of right-wing extremists on TikTok.

Before starting the programme, what was your knowledge of the challenge and its scope?

We were inspired by the movement on TikTok, but when we saw about the challenge from AI4Democracy "Levelling the playing field for political processes", we spent three weeks of market research on the topic prior to the start of the programme to be ready for it.

Why did you choose this particular challenge??

Because of he current relevance of the issue, while democratic actors and parties are so left behind on TikTok, while right-wing extremists can indoctrinate young people unchallenged on this platform.

What progress  have you made from idea to project and pitch? 

We started with the idea of gaining insights into the right-wing extremist bubble on TikTok by using personalised bots and Large Language models to analyse right-wing content.

We progressed  pretty fast from a discussion on architecture, to distributing tasks in the team, and we then moved to development. After four intensive weeks, we are pleased with the MVP we built, as it directly reflects our initial idea and goal.

How do you think AI influences your project?

LLM's and various NLP and ML techniques made our analysis possible; without them our MVP could not have been built.

What's your vision for the next six months? Where do you expect your project/product to be?

Finding a way to continue this project by finding funding options, so we are really focus on that.

What's your personal view on using AI to generate impact for people and the planet?

I think AI offers some unique opportunities for humanity by solving unsolved problems, for example it may enhance biotech research, and it could also boost productivity for society.

However, of course there are many big risks involved. We need to ask who will be profiting from these productivity boosts? Is it society or only a small number of people, with other humans left behind? We need to find political answers to this, like a better distribution of wealth or maybe something like "robot-taxes".

Also, AI’s energy consumption and carbon footprint are massive - this is currently not being addressed, even though we have a climate crisis.

I also feel safety is not taken seriously enough. As our project shows, GenAI can be used to disrupt social media and democracy. In the long run, the alignment of models with society’s wider values and goals needs to be central, but currently, those considerations are deprioritised for profit.

So overall, it's important we should seek ways to use AI  for positive impact wherever we can!