[ryanlchan]

I'm betting it's about talent acquisition.

Like you suggest, the technical part just isn't quite there. I used Summly for almost 3 months, giving it chance and chance again to impress me. It never did. The summaries were mediocre at best and commonly irrelevant, and it got deleted soon after. It's stuck at the point voice recognition is, where it works ok at time, but generally isn't worth the hassle.

Yahoo's been buying up start-ups with good talent but middling traction to do a turn-around. Stamped, Jybe, and now Summly all fit the mold. Yes, they're purchasing at a premium - it's hard to convince a young, scrappy start up founder to join a company with Yahoo's stodgy reputation. But that's exactly what they need - a new infusion of big thinkers to turn around the company.

[robryan]

Having briefly touched this area myself I like to think it is possible to deliver quality summarizations. As with voice recognition the problem is that there is far too much variation in the real world, too many different different standard forms of sentence/ paragraph structure.