Gustav Klimt’s 1900 painting “Philosophy” might have been remembered as a pivotal artwork. Made at a turning point in the artist’s career, it was vividly colored, dramatically composed — even provocative in its blatant nudity and unflinching emotion. But in 1945, the work was destroyed in a fire and essentially lost to history.

For decades, only black-and-white photographs of “Philosophy” existed. Now, thanks to artificial intelligence, we can see the work in full color. But does the re-creation really look like the original? Does it even look like a Klimt?

The new version, created by Google Arts and Culture using machine learning, shows a very different Klimt than you’d expect if you’re familiar with “The Kiss” or “Portrait of Adele Bloch-Bauer I.” On the left side of the canvas, interwoven, nude bodies create a fleshy, blue-hued form that looks almost bruised. On the right, an expressionless, algae-colored sphinx dominates a lurid green sky. A glance at the colors in the reconstructed work, and you might mistake Klimt for his Fauvist contemporaries or think he was hanging around with the painter Marc Chagall.

“Philosophy” is one of three massive “faculty paintings” that the Austrian painter (1862-1918) created for the University of Vienna’s assembly hall ceiling and that were lost in a fire at the end of World War II. Earlier this year, Franz Smola, a curator at Vienna’s Belvedere Museum, which has the largest Klimt collection in the world, guided the paintings’ “recoloring” for the Google Arts and Culture “Klimt vs. Klimt” exhibition online.

“I don’t know any better than Google what those paintings really look like, but I don’t think that they looked like that,” says Jane Kallir, longtime director of the Galerie St. Etienne in New York, which gave Klimt his first shows in the United States. “These things look like cartoons. They don’t look like Klimt paintings.

“It’s like people who try to clone their dogs. You can do it, but it’s not the same dog.”

The faculty paintings are one of several recent attempts to use artificial intelligence to re-create lost art. The Rijksmuseum in Amsterdam used AI to reconstruct missing panels from the edges of Rembrandt’s famous “Night Watch” and, over the summer, temporarily installed them alongside the real thing. A pair of researchers in the United Kingdom, who call themselves Oxia Palus, say they’ve rebuilt a Picasso nude that was hidden beneath “The Blind Man’s Meal,” using 3-D printing and AI. In October, an orchestra in Bonn, Germany, “played” Beethoven’s 10th and unfinished symphony in full. The version was written by an algorithm.

George Cann, co-founder of Oxia Palus, posits that artificial intelligence “could give us this parallel alternative universe of art that we never really quite had.”

It’s an alluring idea. Peek beneath a Picasso at an earlier painting under the surface layer and it’s like you’re peering into the artist’s mind, eavesdropping on thoughts from a century ago. See a painting that was lost to catastrophe come back to life and it’s like you’ve traveled back in time, reversed fate. But if any of this re-created universe of lost art, like “Philosophy,” is inaccurate, the AI creators might not be resurrecting history but inadvertently rewriting it.

Klimt’s faculty paintings make a particularly compelling case for the kind of rediscovery promised by AI.

In 1894, Austria’s Ministry of Education commissioned Klimt to paint allegories representing the disciplines of Medicine, Philosophy and Jurisprudence, expecting he’d use the same traditional mural style that he was known for around Vienna. That’s not what they received. Klimt painted chaotic, dark images of suffering. A sick-looking, motley crew appears in “Medicine,” and “Jurisprudence,” which shows a gaunt man beneath three glamorous, gold-clad judges, seems to depict not the law’s strengths but its elitism.

The works were lambasted by academics who suspected Klimt was mocking their disciplines and by conservative politicians who believed the depictions of female sexuality and nudity would incite immoral behavior. Klimt bought back the paintings from the state and never again did a public commission.

It was a watershed moment in the artist’s career. With the faculty paintings, Smola says, Klimt “started for the first time to do his own thing, to go his own way,” adding that, had they survived, they would be displayed in museums as landmarks of European Symbolist art.

That was part of the impetus to recolor the works with AI. Emil Wallner, a researcher at Google, built the algorithm. He used 100,000 art historical references and programmed it to have a bias toward Klimt’s style. For his part, Smola combed through articles and texts where writers responded to the faculty paintings, seeking an objective sense of color in the subjective writing of criticism.

Throughout the process, Smola had to make bold choices, such as instructing the AI to color “Jurisprudence’s” background red, a decision made after discovering that one of Klimt’s biggest critics noted that “Jurisprudence” featured the colors of the German flag.

Wallner says that in the recolored paintings, he saw a new, “rebellious” side of Klimt. “When you see his most famous works like “The Kiss” and the gold [period] artworks, it can be easy to forget his spirit and who he was as a person,” he says.

But for Kallir, there is little of Klimt in what she calls the “gaudy” re-creations, adding that the paintings would have been more subdued, with smoother transitions from one color to the next.

“If you’ve got a decent eye, and you look at the black-and-white reproductions and compare them to other paintings that were done around the same time, you can probably get a better idea of what they really look like,” she says.

Each application of AI to art has different levels of rigor. To reconstruct Rembrandt’s “Night Watch,” Rob Erdmann, senior scientist at the Rijksmuseum, collected 55 terabytes of data, collaborated with the museum’s art historians, and used a copy of the original painting as a reference. All to give visitors, he says, “the sense — if they squinted — of what the [‘Night Watch’] might have looked like if it had not been cut down.”

But terms like “digital restoration” can confer a misleading legitimacy. To make a 3-D re-creation of the Picasso underpainting, Oxia Palus used an X-ray available online and guided its algorithm with paintings from the artist’s broad “Blue Period.” They didn’t talk to art historians about the original work, but told CNN that “the treasure [Picasso has] hidden for future generations is finally being revealed.”

Kenneth Brummel, who co-curated the new exhibition “Picasso: Painting the Blue Period” at the Art Gallery of Ontario, has reservations about Oxia Palus’s approach, noting that to re-create a canvas’s surface, they would need information accessible only to the museum that owns the work.

“My concern about the methods used here is [not just] that there is incomplete data,” Brummel says, “but also that the algorithms that are used are based on a group of individuals’ arbitrary and highly subjective selection of works of art that they deem to be related.”

Brummel’s exhibition explores technical methods used to learn more about Picasso’s underpainting, including advanced microscopy and spectroscopic imaging. He says he prefers to present the raw scientific data because it allows visitors to draw their own conclusions.

“Part of the beauty of providing a narrative that is incomplete is that you’re inviting others to participate,” he says.

From one perspective, mechanically reproducing lost art does the opposite: It offers a clean answer where there are none, and relies on the predictable when the beauty is, often, in the unexpected.

“Whether it’s an artist or a composer or whatever, there is such a thing as genius,” Kallir says. “The reason that we are awestruck by a Beethoven symphony or a Klimt painting is because they had something that’s inimitable.”