Jim Acosta sparks fury with ‘interview’ of dead Parkland teen’s AI avatar | The video adds to the growing list of AI-video resurrections that people have called “unsettling” and “grotesque.”
https://www.washingtonpost.com/technology/2025/08/05/jim-acosta-joaquin-oliver-parkland-ai/
7 Comments
“Jim Acosta sparked an intense backlash this week with a video of him talking with an AI-generated avatar modeled on Joaquin Oliver, a teenager killed in the Parkland high school shooting in 2018.
Trained on an old photo and audio recordings of Oliver, the AI avatar used a chatbot to generate answers and delivered them in what sounded like his voice.
During Acosta’s conversation with the computer program, he celebrated it as “so insightful” and a “beautiful thing,” saying, “I really felt like I was speaking with Joaquin.”
The video was panned online as “extremely unsettling” and “ghoulish,” with many people citing concerns that such technology could be used to create beliefs the person may not have supported and to tarnish the memory of the dead.”
Who the fuck thought this was a good idea? Let’s see what some kids who were murdered violently at school think, that won’t be weird right.
I want to play chess with Albert Einstein. I think it would be really neat to have idle conversation while playing, taunting him, etc.
I want to have a fireside chat with Abraham Lincoln too. Listening to his voice and seeing his rough personification react to my questions.
If the likeness of famous people, who are no-longer living, can be ‘simulated’… what harm is there in bringing them into a digital livelyhood?
Is there some legal or moral obligation for us to have a persons permission to digitally replicate them beforehand? How about for ‘educational’ purposes?
If computerized technology can scan videos of me, examine the audio and identify my personality, and then generate a digital representation… that is a wonderful thing.
Sign me up! I would be honored to continue digitally interacting with society and my next of kin after my body stops working.
>…people have called “unsettling” and “grotesque”
I’d add “bullshit” to that list.
How does this have any journalistic credence? He did not interview [the person] he interviewed a computer’s estimation of the totality of that person. They’re not equivalent. The AI is limited to the data that was provided to it; an actual person can think beyond the boundaries of their experience. This is true of anything involving these AI “reconstructions”.
AI has just been a boon for sociopaths and attention whores.
I find this reprehensible but there will be business built on it. Maybe for prison punishment with your loved ones, who knows.
At least it highlights how ‘no regulation for 10 years’ will impact us negatively. There won’t be enough subscription money in keeping GMA AI on the mantle after all.