Yes, making RDF popular and widespread is really what Tim had in mind when he promotes the semantic web. Its not about users; its about machines understanding (in the AI sense) each other's content. Sorry to hear that in reality RDF isn't there yet. Tim's vision is that someday soon database users would push the RDF button on their web database, and it would make available their data for other machines (not necessarily people) to use. Specifically he made the point that the semantic web is not about manually adding meaning-tags to html. Its about automation.
Also interesting was the next question at the session, (I was there) by Andrew McCallum, who commented about the semantic web's inability to express a fact's degree of certainty. Andrew is from the probabilistic AI tribe, and rightly characterized the work done to date on the semantic web as coming from old-school AI. This limitation will probably do more to hamper its development in the long run than the ones idenfitified by Novig.
>> a rapidly serviceable, man-rated heavy launch vehicle How about a rapidly serviceable, man-rated _light_ launch vehicle? Its arguably a mistake to have combined the freight transport needs in the inappropriately named Shuttle with the passenger "shuttle" vehicle. So NASA would be better off having a separate human passenger vehicle solely for transporting humans, perhaps something that gets launched from an airborne platform, like the old X-nn concepts. Then NASA avoids the expense and demand of making a 140 ton vehicle man safe (a goal that apparently has not been met). What would we do with the current shuttle craft? Modify them to fly as un-personed freight transports.
. The important thing *is* to realise your model is imperfect; to allow (as well as you can) for that uncertainty in your model; to explore different ways of setting up your model; and then to use statistical inference (Bayes theorem etc) to improve your predictive model as the data comes in.
There is a renegade group of AI folks who study uncertainty -- see http://www.auai.org. Its curious how an entire sub-field can be overlooked in this general discussion.
Actually, the decision-theoretic approach that much uncertainty research is based on (Bayes methods included) stands in contrast to the cognitive approaches in the AI mainstream. The term "statistical" carries too much baggage, some of which litters this thread. For fascinating insight into the rough edges of AI and science generally, look into the controversies between classical statistics, Bayesian, logicalist and other approaches to reasoning.
You seem to have overlooked the fact that Yahoo's search engine is Google, and that generally, Yahoo has overtaken other portals in content, as ugly as that that may appear. Such rampant speculation.
I've been working for our head of software architecture, who jokingly clled himself "junior CTO", and with the departure of the founding CTO will be filling his shoes. Both of them are founders, and kept the flame alive about what the key technology in the product is, and, more importantly what it should be. Neither of them were management positions, nor were they under the VP of Engineering (or vice versa). As we grow (we are currently ~ 250 world wide, with 80 in Platform Engineering,) the CTO function looks like the replacement for the traditional research/ labs role. (Setting up a corporate lab often failed because the researchers were isolated and felt no allegiance to the company's product goals.)
The CTO is often the tag given to the inventor/ PhD in a venture. The CTO function often interacts significantly with customers -- just as the CEO might be the super-sales person in a startup, the CTO is the super-sales engineer.
Of course, if a company does not recognize the need for technology drive behind the product, or the technology in the product is "thin" (like many recent internet plays), then the CTO role is trivial.
In a 'typical' startup the founding folks, ranked by importance (i.e. fraction of shares) may go like this:
- President (may also be head of sales or marketing before those are distinguished.)
- VP Finance (or CFO - may not be necessary until the company has plans to go public.)
- VP Product Development (e.g. Engineering)
- VP Operations (also may be head of IT, Tech. support, etc.)
- CTO
Add to that exposure to _non_ computer languages (like spoken languages:) He (she?) will soon lose a child's magical ability to pick up foreign languages by effortless mimicry. (and you might have to travel abroad with the child to have the same effect on them.:)
More seriously -- pure math/ bi-lingual ability/ ability to read and write music -- all seem to touch on a common fundamental mental skill.
Actually this sounds pretty practical. I'd use something like the Nuance speech reco engine, which has a grammar based recognition step that constrains the text conversion. The grammars are word based regular expressions (and a bit more.)Conceivably one could write re grammars for fragments of the language that you are programming in. You'd have to write these specialized grammars, though.
Global cellphone access might be an idea whose time hasn't come, but consider an Iridium uplink as an alternative on commercial aircraft to flight recorders. (granted you'd still need recorders over the poles) I believe the baudrate of Iridium actually exceeds flight recorder data rates. Planes could transmit airplane flight characteristics in real-time for ground-based analysis. But more interesting is the possibility of assistance to aircraft in trouble as events are unfolding...
Slashdot Mirror
Yes, making RDF popular and widespread is really what Tim had in mind when he promotes the semantic web. Its not about users; its about machines understanding (in the AI sense) each other's content. Sorry to hear that in reality RDF isn't there yet. Tim's vision is that someday soon database users would push the RDF button on their web database, and it would make available their data for other machines (not necessarily people) to use. Specifically he made the point that the semantic web is not about manually adding meaning-tags to html. Its about automation.
Also interesting was the next question at the session, (I was there) by Andrew McCallum, who commented about the semantic web's inability to express a fact's degree of certainty. Andrew is from the probabilistic AI tribe, and rightly characterized the work done to date on the semantic web as coming from old-school AI. This limitation will probably do more to hamper its development in the long run than the ones idenfitified by Novig.
>> a rapidly serviceable, man-rated heavy launch vehicle
How about a rapidly serviceable, man-rated _light_ launch vehicle? Its arguably a mistake to have combined the freight transport needs in the inappropriately named Shuttle with the passenger "shuttle" vehicle. So NASA would be better off having a separate human passenger vehicle solely for transporting humans, perhaps something that gets launched from an airborne platform, like the old X-nn concepts. Then NASA avoids the expense and demand of making a 140 ton vehicle man safe (a goal that apparently has not been met). What would we do with the current shuttle craft? Modify them to fly as un-personed freight transports.
There is a renegade group of AI folks who study uncertainty -- see http://www.auai.org. Its curious how an entire sub-field can be overlooked in this general discussion.
Actually, the decision-theoretic approach that much uncertainty research is based on (Bayes methods included) stands in contrast to the cognitive approaches in the AI mainstream. The term "statistical" carries too much baggage, some of which litters this thread. For fascinating insight into the rough edges of AI and science generally, look into the controversies between classical statistics, Bayesian, logicalist and other approaches to reasoning.
You seem to have overlooked the fact that Yahoo's search engine is Google, and that generally, Yahoo has overtaken other portals in content, as ugly as that that may appear. Such rampant speculation.
I've been working for our head of software architecture, who jokingly clled himself "junior CTO", and with the departure of the founding CTO will be filling his shoes. Both of them are founders, and kept the flame alive about what the key technology in the product is, and, more importantly what it should be. Neither of them were management positions, nor were they under the VP of Engineering (or vice versa). As we grow (we are currently ~ 250 world wide, with 80 in Platform Engineering,) the CTO function looks like the replacement for the traditional research/ labs role. (Setting up a corporate lab often failed because the researchers were isolated and felt no allegiance to the company's product goals.)
The CTO is often the tag given to the inventor/ PhD in a venture. The CTO function often interacts significantly with customers -- just as the CEO might be the super-sales person in a startup, the CTO is the super-sales engineer.
Of course, if a company does not recognize the need for technology drive behind the product, or the technology in the product is "thin" (like many recent internet plays), then the CTO role is trivial.
In a 'typical' startup the founding folks, ranked by importance (i.e. fraction of shares) may go like this:
- President (may also be head of sales or marketing before those are distinguished.)
- VP Finance (or CFO - may not be necessary until the company has plans to go public.)
- VP Product Development (e.g. Engineering)
- VP Operations (also may be head of IT, Tech. support, etc.)
- CTO
John Mark Agosta
Edify Corp.
Add to that exposure to _non_ computer languages (like spoken languages :) He (she?) will soon lose a child's magical ability to pick up foreign languages by effortless mimicry. (and you might have to travel abroad with the child to have the same effect on them. :)
More seriously -- pure math/ bi-lingual ability/ ability to read and write music -- all seem to touch on a common fundamental mental skill.
-
Actually this sounds pretty practical. I'd use something like the Nuance speech reco engine, which has a grammar based recognition step that constrains the text conversion. The grammars are word based regular expressions (and a bit more.)Conceivably one could write re grammars for fragments of the language that you are programming in. You'd have to write these specialized grammars, though.
Global cellphone access might be an idea whose time hasn't come, but consider an Iridium uplink as an alternative on commercial aircraft to flight recorders. (granted you'd still need recorders over the poles) I believe the baudrate of Iridium actually exceeds flight recorder data rates. Planes could transmit airplane flight characteristics in real-time for ground-based analysis. But more interesting is the possibility of assistance to aircraft in trouble as events are unfolding...