My post in the main thread on giantbomb.com enumerates the real technical problems highlighted in Vinny's session. They're manageable. I feel that Vinny did make the game look terrible, to put on a good show. Using keyboard correction, for example, would have solved 90% of the heartburn, but then he wouldn't have such a juicy show, right? He only used it once or twice at the end. Also, he asked for the robot's commands, but never used the information. Why? To issue unsupported commands?
_________________________
First off, my apologies for any GB users that act like dicks towards you. This is an online community, so there will always be those people. I know that the GB crew doesn't appreciate those kinds of people, but they are still here.
Ok, so I replied to your post in the comments section after you posted, but I will re-post it here (slightly modified):
Out of curiosity, how much training is needed to have the experience in the videos you linked? I mean, Vinny went through a good 15-20 minutes of it. Do you have to repeat the training 2-3 times for it to work properly? As others have said, Vinny was very generous in trying to speak as clearly as possible, often directly facing the microphone (and yes, a boom mic should work fine for something like this as long as there isn't background noise, which there wasn't).
I respect what you guys are doing, but when there is voice recognition software out there (Siri, Google Voice, etc) that accurately recognizes voice and context with zero training, I can completely see why someone would become frustrated with this game, as they would expect the voice recognition to be as a good as tech that has been existing for years. If "old" technology does all that with zero training, it is expected that new technology, with training, should be at least up to par with what currently exists. Yes, I realize that an independent developer isn't Google or Apple, but most people don't consider that when it comes to expectations. That is something youshould consider.
I should also mention that, even when Vinny's commands were recognized, or when he typed commands, the robot often still didn't understand, or performed an action totally different than what was commanded. I mean, just watch the last 15 minutes of the video. The voice recognition wasn't the main issue at all. There were several points where Vinny manually corrected the recognized text by typing (as you suggested), and the robot still didn't follow the commands properly. The whole pepper grinder situation was laughable, not because of a lack of voice recognition, but because of the robot did not do what was commanded. When you tell the robot to put something down, then click directly on where he should put it down, he shouldn't then turn around and put the object down on a completely different surface. That's not Vinny's issue, that's a game issue.
I wish you the best of luck with this game, and I honestly hope that you respond to Patrick when he gets in touch with you. I would love to see how the game performs with a fully "trained voice profile", but I have a feeling much of the frustrations will still be there since Vinny's attempts at manually entering commands still proved fruitless. As someone who used to review games (TouchGen.com) for several years before the site closed, I know first-hand that how developers think players will play their games, and how players actually play their games are two completely different things.
Log in to comment