Hwang Woo-Suk is the bad boy of genetics. He's most famous for falsely claiming to have cloned human stem cells. This is, you can imagine, very bad in science. Yet last week, the South Korean researcher was in Siberia, drilling cells from the bones of a 28,000 year-old frozen wooly mammoth. The bones are the only place Hwang is going to find the DNA he needs to bring a mammoth back to life.

Hwang's plan is probably doomed. Not because it's impossible—plenty of the world's best biologists are convinced that cloning a mammoth is just a matter of putting the right minds to work with the right technologies. Even most critics are more concerned with "should we" than "could we" at this point. The problem with Hwang's particular approach is it requires an intact strand of mammoth DNA. And every single strand of DNA in every single frozen mammoth carcass was almost certainly wrecked thousands of years ago by waves of cosmic radiation.

Cosmic energy is trying to rip apart your DNA, too. Also your dog, that tree, the mites on your face, and the all the elephants still alive in Africa. They're all under constant genetic assault from ultraviolet energy. But with life comes enzymes that constantly scuttle up and down your body's trillions of double helices, proofreading them for errors. "The reason you can have 1000-year-old plants is because they are repairing their DNA every single day," says George Church, a geneticist at Harvard who is also working on reviving the mammoth with the Long Now Foundation. "The problem with mammoths isn't freezing. The problem is ionizing cosmic radiation. Some of the samples are perfectly frozen—they stay frozen—but the DNA is completely trashed, and I've never seen researchers address that."

Beth Shapiro, an ancient DNA expert at UC Santa Cruz, agrees. She says a colleague of hers was with Hwang the last time he drilled cells out of a mammoth bone. The mammoth sample looked like it had intact cells, but the DNA was shredded into strands about 600 base pairs long. "This is incredibly well-preserved," she says. "The average ancient DNA is about 100 base pairs."

Church says the right way to do it is by using a relatively new technology called CRISPR/Cas9, a sort of cut-and-paste tool that lets molecular biologists move around bits of genetic material. "Even if the sample is fragmented, we can assemble it in a computer then compare to Asian elephant its closest relative," says Church.

The hitch is that CRISPR/Cas9 is really expensive. The mammoth genome is about 4 billion base pairs long, and piecing the thing together 600 bits at a time would be a mastodon of a task. Which is probably one of the main reasons why Hwang is still drilling for that big score. Of course, we can't be sure if that's exactly the route that Hwang's taking—he's been understandably media shy since it came to light that he had falsified data, embezzled research money, and coerced his assistants into donating their eggs for his cloning projects. I contacted Hwang's assistant and the Yakutsk Mammoth Museum, where Hwang is doing his work, but my questions about his mammoth cloning methodology went unanswered. But those familiar with his efforts, like Shapiro, say Hwang is still trying to find that intact genome. And despite the fact that Shapiro and Church think his assumptions are flawed, they both believe mammoth cloning is going to happen. Shapiro is even writing a book about it.

Not everyone's outlook is so rosy. "Synthesizing a full mammoth set of chromosomes is not, to my knowledge, something that currently could be done very easily, to put it mildly," says Paul Knoepfler, a geneticist at UC Davis. So far, he says, science has only been able to create synthetic yeasts.