Horrific news floated on social media when Elon Musk’s X platform promoted news about Iran attacking Israel to the front page of its news feed, Explore.
On Thursday, one of the titles on X’s main news feed read,
“Iran Strikes Tel Aviv with Heavy Missiles,”
This would have been a rather worrying update considering the geopolitical scenario and recent Israel’s attack on Iran’s embassy in Syria, which killed top officials, and the tensions between the two nations, which escalated further during Israel’s Gaza invasion. So not a good time perhaps, but an ideal time may be for those who want to promote misleading headlines for whatever reasons, and the role of AI (artificial intelligence) cannot be ignored when it can be used as a tool for such heinous acts.
The headline was fake, and more worryingly, it was generated by the X’s official AI chatbot Grok and then promoted to the front page of X’s Explore trending news on the day when this feature was updated to a new version.
Elon Musk’s automation drive – Why the incident took place in the first place
Tracing back to the incident’s roots, it becomes obvious that the Grok-generated misleading title came out as a product of the transition of critical content curation work handled by humans now handed over to artificial intelligence, which yet is not completely intelligent.
The X platform previously Twitter had a feature that provided written context for trending news topics. Probably those descriptions were human-written. Twitter’s algorithm was able to catch relevant keywords and share the trend with trending topics based on the keyword suggestions.
Later, Twitter formed a team of human editors to curate news and manage trends for better results, so the trending topics would no longer be left at the mercy of algorithms to pick; they rather had a human element behind them, so a more reliable approach was adopted. Twitter also partnered with top news agencies like Reuters and AP to further strengthen the process of human-generated context for the trending topics.
But things changed when Musk acquired Twitter in 2022, rebranded it as X, and replaced the human-written context on the trending news. The entire team of human editors was laid off as Musk preferred AI to curate news more efficiently, but it had its own flaws.
The risks of relying solely on AI
The fact that Musk is so ambitious for Grok and promotes it on any occasion, despite knowing that it is still in its early phases and not that reliable, is a concern in its own right. Musk replacing human editors with AI may have some financial benefits, but it comes with its own costs as well.
The fact that the headline was not taken from someone’s feed but rather generated by Grok, the native AI chatbot on the X platform, But the most concerning fact was that even if Grok generated such a false headline in the first place, it was then promoted to the top of the Explore news page.
This shows that there was not a check system in between the generation and promotion of this news, and if there was any system, it wasn’t sufficient enough, and the company totally relied on AI to do the job. As said earlier, AI is still not smart enough to handle sensitive topics, despite the fact that it offers many benefits of dissecting data and provides results with more efficiency than humanly possible. Still, there are areas that need human intervention, and AI may take some time to develop to that stage. Until then, relying solely on AI procedures is not a smart move.
Original story at Mashable
