By TUDOR FINNERAN
Global Business Journalism reporter
In the rapidly evolving intersection of technology and journalism, Ali Tehrani, co-creator of Newswyze and Gist, stands as a prominent figure leveraging artificial intelligence to reshape the landscape of business and media. The Iranian-born journalist’s expertise in natural language processing has led him to collaborate with the International Center for Journalists, where he imparts his knowledge on cutting-edge technology advance.
Recently, Tehrani took center stage as a guest lecturer addressing Global Business Journalism students at Tsinghua University. Providing a unique perspective on the challenges and opportunities posed by AI, he went onto explain the importance of prompt-based language changes, how political ideologies will influence results and the importance of geographic differences in AI.
In a virtual setting, characteristic of a technologist and innovator, Tehrani emphasized the importance of education about the latest technology in the work place. During this new age of AI, “educating journalists who often have extreme [negative] views on AI” is priority number one, he told the Global Business Journalism students via Zoom.
Tehrani emphasized the influence of geographic location and political agendas on AI systems, cautioning these very systems often produce varying results based on their “training” data. While he underscored the efficiency of AI in tasks such as data analysis, coding, and image creation, he also pointed out a critical flaw – the lack of a mechanism to balance the inherent biases within AI models.
Because the creator of the AI program controls the information that will be included in it (known as “training the data”), AI products are prone to political bias, propaganda, misinformation and intentional falsehoods.
“Authoritarian regimes with high propaganda likely have higher risk of propaganda-influenced datasets,” he said.
This deficiency, he argued, could lead to the amplification of certain opinions and false information, posing a significant challenge for the efficacy of AI. Tehrani went onto explain how this is especially true in the case of “filter bubbles on social media platforms” which far too often lead to “the reinforcement of extreme views.”
Ali Tehrani: Financial incentives exist to spread disinformation
Delving deeper, Tehrani addressed the issue of disinformation and propaganda stemming from geographic data pools with a high concentration of misleading or false content. AI experts and natural language processors, he asserted, are actively working to mitigate these challenges. He identified one root cause for false and misleading content as financial (rather than political) incentives. Tehrani said predictive algorithms tailor content based on a user’s digital habits and those business models are likely to be manipulated for AI to show content skewed toward the user’s ideological preferences.
Tehrani also illuminated the potential benefits of integrating AI with journalism. He argued that, if reliability and verification challenges can be addressed, the industry could witness a new era of expedited journalism, minimizing repetitive tasks and streamlining data analysis. This is reliant on “the importance of journalist engagement,” he said, specifically relating to “shaping the technology to align with journalistic values.”
Tehrani suggested that the industry stands on the cusp of a transformative era. Speed, reliability, and added value are paramount, promising not only a revolution in reporting but also in service provision, monetization, and overall business models, he said.