Introduction

In the context of rising disinformation in the lead-up to the U.S. 2024 presidential election, the urgency to combat online falsehoods has never been greater. Disinformation has not only swayed public opinions about political candidates but has also seeped into critical discussions around public health, climate change, and immigration. As a grim addition to the fight against misinformation, generative AI tools can produce deepfakes and concoct fictional facts at unprecedented speed, worsening this pervasive issue.

The Emergence of Factiverse

Amidst this chaos, Norwegian startup Factiverse has emerged as a beacon of hope. This groundbreaking company recently showcased its potential at the TechCrunch Disrupt Battlefield 200 event in October, where they clinched the award for best pitch in the Security, Privacy, and Social Networking category. Factiverse has developed a cutting-edge business-to-business tool that offers live fact-checking services across text, video, and audio formats. Their mission is clear: to help companies save valuable hours in research while shielding them from reputational damage and possible legal repercussions.

Founding and Funding

Founded in 2020, Factiverse may still be in its early days, but it has already raised approximately $1.45 million in pre-seed funding. The company is not just sitting idle; it has initiated collaborations with a variety of media entities and financial institutions, including one of Norway's largest banks, according to CEO and co-founder Maria Amelie.

Fact-Checking During U.S. Presidential Debates

During the recent U.S. presidential debates, Factiverse played a pivotal role by providing live fact-checking services that many media outlets utilized. "We’re not just another LLM (large language model). Our system is built on a distinct model based on information retrieval," Amelie explained.

Founders’ Background and Vision

Bringing a wealth of experience from her background as a technology journalist and published author, Amelie co-founded Factiverse alongside Vinay Setty, who serves as the company's CTO and is also an associate professor in machine learning at the University of Stavanger. Their aim was to craft a solution focused specifically on B2B applications.

Training Methodology and Data Integrity

Factiverse is particularly proud of its training methodology: the AI model relies on rigorously curated, dependable data from credible sources and established fact-checkers. This sets it apart from the 'junk food data' typically associated with generative AI. “We train our AI to understand and think like an experienced researcher, ensuring top-notch verification,” Amelie stated.

Technological Capabilities

Harnessing the powers of machine learning and natural language processing, Factiverse’s model can identify potential claims and search the web in real-time, spanning major search engines like Google and Bing as well as specialized platforms like You.com and various academic papers. “What makes our tool exciting is that we don’t just pull whatever ranks highest in search results. Instead, we recommend the most historically credible sources for your topic,” Amelie noted.

Competitive Edge

Factiverse is breaking new ground by transcending what many traditional AI models can offer. Currently, they report that their technology surpasses GPT-4, Mistral 7-b, and GPT-3 in identifying claims worthy of fact-checking across 114 languages. The effectiveness of their model in determining the accuracy of claims currently stands around an impressive 80%.

Future Aspirations

With aspirations to solidify its presence in the U.S. market and a goal to begin a seed funding round by 2025, Amelie is optimistic about the company’s future. “We have the resources to excel, but our ambition is to be the quickest to adapt and grow. We are actively seeking clients and investors interested in investing in trust and credibility,” she revealed.

Conclusion

As disinformation continues its rampant growth online, startups like Factiverse are leading the charge towards a more informed public, where truth and reliability take precedence over falsehoods and sensationalism.