Social Media Buzzes Over Awkward Stage Interaction between AI CEOs and PM Modi
Sam Altman and Dario Amodei avoid holding hands with PM Modi on stage. Awkward moments go viral.
An awkward on-stage moment at the India AI Summit as OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei stand beside PM Narendra Modi during a group photograph, highlighting visible tensions within the AI industry.
The India AI Summit commenced in Bharat Mandapam with a grand inaugural ceremony.
However, something went awkwardly at the India AI Summit stage this morning. The CEOs of
rival AI firms OpenAI and Anthropic refused to hold hands together in order to capture a group
photograph. The awkwardness was caught on camera and speaks about the AI industry
tensions. It has become a subject of jokes on social media.
Here’s what actually happened at the AI summit
After the important address at the India AI Impact Summit, some of the VIP tech leaders got
onstage for a group photograph with the PM. The group consisted of Google CEO Sundar
Pichai, Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, and Meta Chief AI Officer
Alexandr Wang, among others.
As PM Modi looked on, the two competitors were standing next to each other. Both appear quick
and uncomfortable. After Modi’s request, they seemed unwilling to join hands. In the end, the
hands that should show the sign of unity raise their fists. This moment shows the AI Cold War
moment.
The rivalry between OpenAI and Anthropic
Now, OpenAI and Anthropic are counted among the most identifiable names in artificial
intelligence. However, their approach to technology is different. OpenAI is popular for moving
quickly and creating AI tools like ChatGPT easily available to the public. On the other hand,
Anthropic claims to place stronger emphasis on safety and careful behaviour with the ‘Claude
chatbot’.
In an interview in 2023, Amodei had declared that he quit OpenAI due to directional differences.
He said, “You needed something in addition to just scaling the models up, which is alignment or
safety. You don't tell the models what their values are just by pouring more compute into them,”

