Group items tagged

Two years ago, the two greatest Jeopardy champions of all time got obliterated by a computer called Watson. It was a great victory for artificial intelligence--the system racked up more than three times the earnings of its next meat-brained competitor. For IBM’s Watson, the successor to Deep Blue, which famously defeated chess champion Gary Kasparov, becoming a Jeopardy champion was a modest proof of concept. The big challenge for Watson, and the goal for IBM, is to adapt the core question-answering technology to more significant domains, like health care. WatsonPaths, IBM’s medical-domain offshoot announced last month, is able to derive medical diagnoses from a description of symptoms. From this chain of evidence, it’s able to present an interactive visualization to doctors, who can interrogate the data, further question the evidence, and better understand the situation. It’s an essential feedback loop used by diagnosticians to help decide which information is extraneous and which is essential, thus making it possible to home in on a most-likely diagnosis. WatsonPaths scours millions of unstructured texts, like medical textbooks, dictionaries, and clinical guidelines, to develop a set of ranked hypotheses. The doctors’ feedback is added back into the brute-force information retrieval capabilities to help further train the system.

For Watson, ingesting all 2.5 million unstructured documents is the easy part. For this, it would extract references to real-world entities, like corporations and people, and start looking for relationships between them, essentially building up context around each entity. This could be connected out to open-entity databases like Freebase, to provide even more context. A journalist might orient the system’s “attention” by indicating which politicians or tax-dodging tycoons might be of most interest. Other texts, like relevant legal codes in the target jurisdiction or news reports mentioning the entities of interest, could also be ingested and parsed. Watson would then draw on its domain-adapted logic to generate evidence, like “IF corporation A is associated with offshore tax-free account B, AND the owner of corporation A is married to an executive of corporation C, THEN add a tiny bit of inference of tax evasion by corporation C.” There would be many of these types of rules, perhaps hundreds, and probably written by the journalists themselves to help the system identify meaningful and newsworthy relationships. Other rules might be garnered from common sense reasoning databases, like MIT’s ConceptNet. At the end of the day (or probably just a few seconds later), Watson would spit out 100 leads for reporters to follow. The first step would be to peer behind those leads to see the relevant evidence, rate its accuracy, and further train the algorithm. Sure, those follow-ups might still take months, but it wouldn’t be hard to beat the 15 months the ICIJ took in its investigation.

I have not created any new way of writing. All I'm doing is writing computer programs that mimic the way people write. Going back to the Elizabethan sonnets, Shakespeare or one of his contemporaries created the 14-line iambic pentameter poem, where the rhyming pattern was 'a-b, a-b, c-d, c-d, e-f, e-f g-g.' G-g being a couplet at the end. By line 9 there has to be a turn in the poem, so there has to be a phrase like 'yet' or 'but.' The first line is typically a question, which acts as a title. All of them are 10 syllables in each line... they have to go in the rhythm of that pattern. If you do an analysis of sonnets, you'll realize that about 10% of sonnets violate those rules. But they do it only in a very particular way. Even that formulation of violation is itself constrained... Once you have all of those rules you then write algorithms that mimic those rules. It's a very different kind of philosophy from artificial intelligence.

The methodologies are extremely old, just like the methodologies of writing haiku poetry are very old. An Elizabethan sonnet is 14 lines - that is a line of code if you think of it that way. The code is constrained. So all genres, no matter what the genres are, are a form of constrained writing.

anticipatory systems. 

Increasingly, rather than waiting for us to tell them what we want, in the form of a search query or command, they'll prompt us with suggestions.

Here's a simple definition of anticipatory systems. Think of them as artificially intelligent services that are aware of external context — including ambient inputs like time of day, social connections, upcoming meetings, local weather, traffic and more.

For the Xbox to be able to turn on, identify who you are, and log in to your profile at a moment's notice--simply at the sound of the voice command "Xbox on"--it needs to do something a bit creepy: It has to be listening to what people are saying in your living room all the time.

Microsoft almost definitely is not recording, let alone uploading and archiving, every sound that happens in your living room 24-7-365. That said, the fact remains that there is a microphone in your home that's always live and connected to some super-smart computing devices--and a very distant server.

Expect Labs made a smartphone app called Mind Meld that demonstrates their listening tech expertise. Mind Meld can listen to the online conversation of a group of people, and detect what they are talking to such a high level of automatic detection of content and context that it can magically suggest online sources of information that might interest the group.

H.G. Wells describing the “world brain” as a  “complete planetary memory for all mankind.”

As a Google engineer told author Nicholas Carr, “We’re not scanning all those books to be read by people. We’re scanning them to be read by our AI.”

social media is becoming a skill, not a job. Companies like Intel and Dell and IBM are leading the way in broadly distributed social participation, giving thousands of employees the opportunity to win hearts and mind in social and with smart content.

This decentralization of social communication has widespread ramifications for social media management software vendors, as it puts additional emphasis on triage and workflow tools.

The days of one social media manager handling Twitter, Facebook, Linkedin and the rest is coming to a close (as is the era of the one or two person content marketing team) and the same way all of us have a corporate email address and phone number, we’ll all (or nearly all) have a role to play on behalf of the company in social and content marketing, eventually.