tag:blogger.com,1999:blog-20402273.post4325498185350044516..comments2024-03-18T06:38:00.224+00:00Comments on Alan Winfield's Web Log: A Crisis of ExpectationsAlan Winfieldhttp://www.blogger.com/profile/08263812573346115168noreply@blogger.comBlogger10125tag:blogger.com,1999:blog-20402273.post-61044941704065797482013-04-06T10:00:30.265+01:002013-04-06T10:00:30.265+01:00Thank you for your very kind comments Joseph - and...Thank you for your very kind comments Joseph - and your blog post.Alan Winfieldhttps://www.blogger.com/profile/08263812573346115168noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-24378418772846516362013-04-06T09:59:14.927+01:002013-04-06T09:59:14.927+01:00Your comments are very much appreciated. Thank you...Your comments are very much appreciated. Thank you!Alan Winfieldhttps://www.blogger.com/profile/08263812573346115168noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-82640829885766177472013-04-06T09:56:23.156+01:002013-04-06T09:56:23.156+01:00Great! Many thanks Bob. I shall read your blog pos...Great! Many thanks Bob. I shall read your blog post with interest, and post a comment there.Alan Winfieldhttps://www.blogger.com/profile/08263812573346115168noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-1378037291589423232013-04-06T09:52:55.874+01:002013-04-06T09:52:55.874+01:00Many thanks Roger. I wasn't aware of the disci...Many thanks Roger. I wasn't aware of the disciplined approach to benchmarking and reference data sets in the speech research community. Very interesting. As you say, that's exactly what we need in robotics. Our benchmarking work in the euRathlon project will I hope contribute to that. Alan Winfieldhttps://www.blogger.com/profile/08263812573346115168noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-45345651056881125012013-04-05T22:29:34.913+01:002013-04-05T22:29:34.913+01:00Hi Alan - thoughtful, persuasive and erudite. Will...Hi Alan - thoughtful, persuasive and erudite. Will chase up on the links as I have much ground to cover. Will start blogging when I have some spare time at http://tinyurl.com/ckbwe6oJosephhttps://www.blogger.com/profile/13060362818013765148noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-64998836913666963762013-04-05T22:28:29.760+01:002013-04-05T22:28:29.760+01:00Hi Alan - inspirational as ever. Will start bloggi...Hi Alan - inspirational as ever. Will start blogging about this thoughtful post - will chase up the links (do email me any additional sources for further reading).<br />http://tinyurl.com/ckbwe6oJosephhttps://www.blogger.com/profile/13060362818013765148noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-46241405989567818392013-04-05T01:05:19.893+01:002013-04-05T01:05:19.893+01:00I'm sorry I didn't read your references fi...I'm sorry I didn't read your references first before commenting. I watched the lighthill report and can understand where your coming from now.<br />Stupid spelling mistakes like "sights" instead of sites don't make me look any brighter either. gullygunyahhttps://www.blogger.com/profile/08188145333933067606noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-66039179719328925192013-03-29T21:53:39.478+00:002013-03-29T21:53:39.478+00:00I've seen robohype and tech hype too. I read a...I've seen robohype and tech hype too. I read a lot of science and tech sights daily.<br />There are dedicated robot news sites but only a few are worth following. My favourites would be... well this is one of them, if only the author published a little more frequently, : D<br />But seriously the ones for reliable daily content in my list are <br />http://spectrum.ieee.org/blog/automaton and robots.net<br />There are others but those 2 I have found keep up with latest developments at the most frequent rate.<br />Hype is good though and winters happen in every place in life.<br />Despite the proliferation of robotic vacuum cleaners, how many actually map out the floorplan and know the most efficient route for cleaning?<br />This hasn't stopped millions of them being sold. Even hobbyists who know this fact(excluding myself) still buy them because they are cool or the nearest thing to a commercial robot they can afford.<br />I think you would be taking on the battle you can never win by trying to eliminate robohype. <br />Hype will always exist as long as humans are willing to deceive for some personal gain. <br /><br /><br />gullygunyahhttps://www.blogger.com/profile/08188145333933067606noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-22278130852358963942013-03-29T13:46:41.685+00:002013-03-29T13:46:41.685+00:00http://sluggish.dyndns.org/blog/?x=entry:entry1303...http://sluggish.dyndns.org/blog/?x=entry:entry130329-153326Bob Mottramhttps://www.blogger.com/profile/07508972554031337434noreply@blogger.comtag:blogger.com,1999:blog-20402273.post-71404417002072327222013-03-29T09:48:25.278+00:002013-03-29T09:48:25.278+00:00Excellent discussion Alan - shame I missed the act...Excellent discussion Alan - shame I missed the actual talk. There's an interesting parallel in my field - speech technology - where we have been through at least three cycles of overpromising followed by funding dips. As most people know, speech technology is now becoming very successful (witness the popularity of Siri) due to the steady technical progress that has been made in the last ~20 years. The main driver has been a very strict evaluation/assessment regime with data sets of increasing complexity being released to the international research community by enlightened funding agencies. Once this regime was put in place, it was no longer possible to make unsubstantiated claims about performance levsl and what thetechnology could actually do. Of course there's a downside - this approach tends to stifle innovation as people become timid about trying new ideas (which inevitably work badly first time), and don't want to risk their own finding. I know that you are leading an evaluation effort in robotics, and this could be key to overcoming the problems you've identified (and could maybe draw some inspiration from what happenend in speech technology).Roger K. Moorehttps://www.blogger.com/profile/06632743126904273531noreply@blogger.com