DAVID VASKEVITCH wants your memories. No, not the RAM chips from your computer, but the ones in your head, whose physical manifestations you spend a lifetime archiving or else leaving to rot in the attic. Love letters, diaries, books, music, sound recordings, emails, lecture notes, movies, photos or camcorder footage 鈥� Vaskevitch foresees computer hard drives becoming so vast, so fast, that we will be able to store a lifetime鈥檚 worth of all these things on a single laptop.
These days, 200-gigabyte disc drives can be bought for well under 拢200. Four years from now drives of 1000 gigabytes (a terabyte) should be commonplace and affordable. 鈥淎ll of a sudden it will really become practical to have all of your life鈥檚 memories and carry them around in a laptop,鈥� says Vaskevitch. 鈥淭hat鈥檚 every picture I might ever want to take, every song I鈥檝e ever heard, every movie I鈥檝e ever seen or might want to see, every book I鈥檝e ever read or might want to read, every note that I鈥檝e ever taken.鈥�
It is tempting to dismiss the idea of such a 鈥渓ife library鈥� as absurd, until you realise that the softly spoken Mr Vaskevitch is the chief technology officer at Microsoft. And whether you see Bill Gates and his crew as arch monopolists or innovators, it鈥檚 hard to dispute the company鈥檚 power to influence the future direction of technology.
Advertisement
Microsoft already has a research team working to make its future PC operating systems capable of not only storing but organising and managing life memories of this kind. But it was a couple of conversations at home that really convinced Vaskevitch that the company was on the right track. When he decided to buy his college-bound daughter a new laptop computer, she simply didn鈥檛 care what type she got. 鈥淪he just said it had to have a really big hard drive,鈥� says Vaskevitch. 鈥淪o I said 鈥極K, how big?鈥�. And I thought she would say 10 gigabytes or 20 gigabytes, but she said 鈥楴o, I want at least 100 gigabytes and, actually, is there such a thing as 1000 gigabytes?'鈥�
The fact that her only worry was having enough room to store digital music and photos struck a chord with Vaskevitch. And when his 75-year-old mother, impressed with his ability to scroll through hundreds of family photos on a slimline laptop, asked for one of her own after having shown no interest in computers, Vaskevitch was certain that Microsoft鈥檚 MyLifeBits project is working along the right lines.
Led by distinguished computer architect Gordon Bell at the company鈥檚 Bay Area Research Center in San Francisco, the MyLifeBits team is working out how future operating systems can usefully organise, catalogue and make searchable the varied types of content, from the scruffiest scanned-in postcard, to newspaper cuttings, to high-definition video. When each digital artefact is stored, it will be annotated with text or speech. To retrieve it, it can be searched for in the same ways. The really tough part, not addressed by any of the current systems, is finding a way to hyperlink related items 鈥� a love letter to a piece of music and a photo, say.
The idea is not to market a self-contained 鈥渓ife database鈥� that people will run like Excel or Power Point. 鈥淒atabases are completely irrelevant to most consumers. They wouldn鈥檛 know what to do with one,鈥� says Vaskevitch. His aim is to integrate MyLifeBits with the operating system, so that the database is invisible but the information is stored in an orderly fashion nonetheless. While he won鈥檛 be drawn on a timescale to launch, he hints that an early take on the technology may form part of the next version of Windows 鈥� currently codenamed Longhorn and due out in 2005.
Before then, the potential to package your life in this way could raise important security and privacy issues, says Ehud Reiter, a computer scientist at the University of Aberdeen in the UK. The police and courts might want access to your memories 鈥� such as letters and photos 鈥� if they are deemed relevant to a legal case. 鈥淚t鈥檚 just one of an enormous number of public policy issues that need addressing,鈥� Reiter says.
Not least among the technical hurdles, says Reiter, will be the development of open data formats that will be readable in future. He thinks a group working along the lines of the World Wide Web Consortium, which has standardised html and other web languages, should be tasked with ensuring our memories are not saved in proprietary formats. He points to the difficulty of playing old 78 rpm records, which were once universal but now require specialist turntables. 鈥淲ho鈥檚 to say Microsoft Word will be readable in 20 years鈥� time?鈥� Reiter asks.
How to build a life library
Microsoft is not alone in wanting to build life libraries that contain many years鈥� worth of data. Last week, the UK Computing Research Committee, a group of computing gurus who define the future direction for the country鈥檚 computing research, unveiled the challenges that must be solved to make the libraries work for real.
鈥� TECHNOLOGY: Memories need to be stored reliably over a period of many decades. So the formats in which they are stored must be non-proprietary so that they can be played on a wide range of devices in future. If a company goes bust, you don鈥檛 want it to take your memories with it.
鈥� SEARCHABILITY: Getting the data in is easy, getting it out is tough. New image-recognition and language-processing techniques will be needed so that people can retrieve not just specific memories but associated ones too 鈥� for example, a high-resolution image taken at the same time as a low-resolution video or all the available images of a particular building.
鈥� SECURITY/PRIVACY: Who has access to a person鈥檚 life library? If someone gets hold of your videos or emails and stores them in digital memory, do you have the right to access this material or edit it? Or any say at all as to how it is used?
鈥� INTERACTION: How will people best view, update, delete and annotate memories? One way might be to allow people to annotate their images with other data such as a voice recording.
