Future of Lazy Artists
David Simmonds

You’ve probably heard about the next big thing in digital intelligence. A computer is now able, for better or worse, to replicate the style of an ancient master painter and create a brand new work that might be from the master himself, if he hadn't been dead for almost 350 years. 

"The Next Rembrandt Project,”

The painting, dubbed "The Next Rembrandt Project,” was commissioned by Dutch banking colossus ING. It`s the signature piece in a new advertising campaign that showcases ING`s sponsorship of the arts, while attempting to stress its focus on innovation. The team behind the painting, led by the J Walter Thompson advertising agency, scanned numerous paintings by Rembrandt. They created algorithms based on the brushstrokes, facial features and background lighting that made Rembrandt`s portraits so distinctive. The researchers fed the algorithms to the computer and told it, essentially, to go to it. The computer produced an eerily familiar looking `original` Rembrandt, which to top it off was produced on a 3-D printer, with a depth of thirteen layers of ink.

Now, if Rembrandt were somehow to wake up, it is probable that he would have a hard time comprehending what a computer is and how it works. The same might be true of certain columnists with the “Times,” too. I think it also likely that Rembrandt would be a little annoyed at how his genius could be reduced to algorithms and then re-applied to produce something `fresh.` He would surely think his portraiture style would belong to him alone, at least to the extent that he would be the only one allowed to use the `Rembrandt` moniker.  

Let`s suppose that he could resolve those concerns and were inclined to coast on his well-earned reputation. Would he not be tempted to forgo the dozens and dozens of hours he would have to commit to creating a new work, in favour of just feeding the parameters into his computer system and telling it to produce a new painting of a man with a hat, a frilly collar and a bulbous nose?

He`d probably earn a decent income from it.

Would he be able to claim it as an original `Rembrandt’ or a `By Rembrandt in the style of Rembrandt,` or a `Rembrandt computer graph’? However it was to be classified, he`d probably earn a decent income from it, although the computer’s programmer might demand a share in the creative rights. So will the only breakthrough of the Next Rembrandt Project be to make successful artists lazy?

Thus, the problems just begin to compound. What`s to stop some fool with an algorithm from flooding the market with a `newly discovered Chef Boyardee Can` series by Andy Warhol, convincing enough to hoodwink the experts? Has anybody given any thought to what will become of art forgers, those loveable rogues who appear destined to patrol the unemployment lines beside middle managers? What about taste: what is to stop the philistine from commissioning an ‘authorized Picasso’ piece depicting him donning the green jacket from the Masters Golf tournament? 

If a computer can now churn out an ‘original’ Rembrandt, will the skill of the artist  now lie not in the application of paint to canvass, but in the selection of styles to copy?  Will an artist now attract attention for originality based upon a formula that is `Sixty per cent Tom Thomson and thirty-five per cent Salvador Dali, with a five per cent hint of Leonardo Da Vinci,’ or ‘One half Alex Colville and one half Marcel Duchamp.’

A big an incursion into the arts.

All this Rembrandt stuff serves to highlight how side of the ledger the computer has made. There have been programs, like ‘Band in a Box,’ through which musicians can program computers to write original tunes. Just last year, Oscar Schwartz gave a TED talk in which he demonstrated that a computer can generate poetry that people think is more human than human poetry; the triumphant computer was raised on Emily Dickinson poems and was going toe-to-toe with Gertrude Stein.’

Is the writing on the wall for the creative arts? No, but I wouldn’t mind getting a hold of those Next Rembrandt algorithms so I can knock off and quickly sell a couple of his ‘newly discovered’ works, just in case I’m wrong.