‘VisionQuest’ Seems Eager To Confront Dangers Of Unregulated AI, But The Team Behind It Says MCU Series Isn’t A Reflection Of Current Events

It’s quite an interesting time for Marvel to release a series about the comic book universe’s sentient AI agents confronting the limits of their programmed humanity, and the series’ stars are not entirely sure what to make of it.

“I think that Marvel has sort of obliquely talked about it since Age of Ultron. They have offered up two sort of versions of what AI can be — a sort of chaotic, psychotic creature in Ultron, who wants to subjugate or even destroy humanity, and then a version in JARVIS and The Vision that wants to sort of augment human beings and believes in human agency and the very thing that AI might see as a flaw, which is inefficiency,” star Paul Bettany tells Deadline. “Humans are very inefficient. Art is inefficient. Music is inefficient. That is part of humans’ beauty. Children don’t become useful until they’re older. I think Vision has been really useful, a tool in that, and revealing what the human condition is.”

However, while the series itself has plenty to say about what’s lost when computers are left to make decisions about the fate of world, Bettany does not.

“I’m not sure we’re the guys to discuss the future of humanity under our AI overlords,” he said.

VisionQuest, which launched earlier this month, sets out to complete Marvel’s WandaVision trilogy by once again picking up in the aftermath of Wanda Maximoff’s (Elizabeth Olsen) hex on Westview and follows White Vision (Paul Bettany) as he tries to reconnect with humanity after his original memories and emotions are restored at the end of the titular 2021 series. Sure, he remembers his life with Wanda and their sons Tommy and Billy, but he doesn’t feel any bond toward that old life of his. Or so he tells himself.

To Bettany’s point, Marvel is bringing back an old villain to force Vision to confront his grief. James Spader‘s Ultron was first introduced in 2015’s Age of Ultron, in which Tony Stark and Bruce Banner create a sentient artificial intelligence model for peacekeeping and it ultimately decides the way to keep the peace is to exterminate the human race. Well, Ultron is now trapped inside Vision’s mind space, where he’s also housing all of the other AIs that Stark Industries has dreamed up over the years including EDITH, FRIDAY, JARVIS, D.U.M.-E. and U.

Spader argues that Ultron is “far more advanced than what we’re seeing, at least in real time,” and the computer program might still have questionable ethics, but he’s had a bit of a change of heart since the last time audiences saw him.

“He’s fascinated with humanity, not fascinated with AI. He’s not looking for a way to advance himself in terms of AI. He’s looking for a way to advance himself within the context of humanity, and that’s telling,” Spader says.

In the fourth episode, which was released earlier this week, Ultron convinces Vision to hand over the keys to the mind space and cede control of his body in a moment of weakness against a new threat: a synthezoid named Jocasta, who is revealed to formerly have been a researcher from a biotech group studying the advancement of sentient AIs.

“I really haven’t thought about this aspect of it, because all of this stuff is happening in our real lives,” Spader continued. “Ever since we did the show, this has been become so accelerating.”

Showrunner Terry Matalas also doesn’t see the series, which shot in the first half of 2025, as necessarily a way to comment on current events but more so broadly float the idea that “AI is our creation. We need to raise it right.”

“Whether that means guardrails, whether that means slowing it down and connecting with it in a different way. So I think that is kind of how VisionQuest touches upon current events in that way,” he explains.

Matalas argues that, fundamentally, the super intelligence that Marvel has concocted is different than what we’re facing in the real world as AI agents are found rooting around in people’s bank accounts without authorization and disabling their own safety monitoring systems during coding runs. Why?

“There’s something about their code, which we’ll learn about in series, that gives them the ability to feel, gives them the ability to connect, and where that comes from… I would say stay tuned,” he said.

Besides Vision, this is the first time that the franchise has bestowed human avatars on the rest of the Stark Industries bots. That has not stopped the computer programs from becoming a beloved part of the MCU, though, which is why the writers can get away with presenting JARVIS deleting his code to clear server space as heroic act of self-sacrifice in the second episode.

“For JARVIS, it was really about — of the AIs in that house, that’s definitely the one Vision is closest with. He’s like a brother. They share so much of the same code, and so it was really telling the audience that we’re playing for keeps here. Nobody is safe,” Matalas weighed in. “It’s a heartbreaker, and we’re sad to see him go, but what does that mean now for the other AIs in that house who are not as in tune with security as he was?”

Since JARVIS is, again, a computer program, there’s really no reason why he can’t resurface in the MCU at some point, especially given that he and Vision share code. It wouldn’t be so dissimilar from SWORD rebuilding Vision in the aftermath of Avengers: Infinity War. Regardless of whether JARVIS is really dead, Emily Hampshire (who plays the tactical bot that Tony Stark built for Peter Parker, EDITH) doubles down that the interesting point is in the human parallels.

“I don’t think [the AIs have] experienced death in that way that it is affecting, and knowing that you can be the one who goes away, I just thought it was such an interesting story point,” she said. “It’s so fascinating and smart of Terry Matalas to put that in there, because that’s what happens when a real family loses somebody. You lose that sense of that security you had with that person and that person’s position in your life, and it just was kind of brilliant.”

So, rather than ruminate on the realities of setting artificial intelligence loose on the world, Hampshire joins her colleagues in hoping that VisionQuest‘s ragtag group of sentient computer programs will teach us all a little more about what it means to be human.

“As humans, we are constantly outsourcing ourselves to tech,” she said. “I think it’s both touching and really funny that these AIs are trying to teach Vision, who is an AI, about humanity, which none of them know anything about — which is, in my opinion, the most human thing ever to pontificate about things you know nothing about.”