Slashdot videos: Now with more Slashdot!
We've improved Slashdot's video section; now you can view our video interviews, product close-ups and site visits with all the usual Slashdot options to comment, share, etc. No more walled garden! It's a work in progress -- we hope you'll check it out (Learn more about the recent updates).
Fit the car with forward-facing LIDAR to scan the road and it can fully gauge potholes a lot better than any driver, give it a twin LIDAR system to compare the reflections from two heights above the road and it can detect water and differing surfaces and suddenly has an amazing amount of information to plan from.
I do think computer controlled cars are going to happen but there are a lot of obstacles to overcome first such as handling non-automated vehicles, the legal framework, and developing 'sufficiently cheap' sensor arrays. I don't expect that difficult road surfaces or conditions to be one of the more difficult ones to solve, though.
The AI doesn't have to be on par with humans in general, just in the specific case of driving- which from what we've seen is a lot simpler for them than software development.
It would have been a lot more profitable to develop a computer which can develop software better than an average developer than one which could beat a grandmaster at chess, and yet what do we have?
The computer will be tracking the motions of pedestrians in case they step into the road, and will likely have noted that many are below statistical average. It will also have access to full demographic information so it'll know exactly what type of neighborhood it's driving through to adopt a driving strategy.
Most importantly the delay between a computer sensing something and reacting makes even the fastest reacting human look ridiculously slow.
I think awful Python is a lot more manageable than awful C++, and I've seen enough of both.
(Probably wrote enough of both too, but I'm too smart to admit that here).
Why does the answer have to be either 'taxi' or 'bus' when it's possible to combine the two and have a sensible multi-drop scheme? This would need a decent number of users (Higher than that needed to sustain a local bus company, I'd guess), but would manage to combine the two nicely, and with computerised routing of vehicles should be practical.
Like a taxi you book from where you are, and it'll come to collect you there and drop you off at your requested destination. Unlike a taxi though there could be others riding already, and the vehicle may divert to make pick ups/drop offs on the way within reason.
The result is something which is similar to a less direct taxi, but will be a a lot cheaper as it's multi-occupant and will have passengers almost continually, similar to a bus. It also means that there are less vehicles on the road as each one is carrying more people, and this reduction would actually improve as the service became more popular- it's easier to have efficient routes when there are more options available. Vehicles would likely range from large cars to small buses, things small enough to get through residential areas (I'm in the UK where getting something the size of a full bus down a side street would be impressive) but large enough to carry a few small groups of people.
If you want to keep the 'bus' mentality too then have scheduled pick-ups over the most-used routes, so 'There will be a vehicle arriving at Easterly School every ten minutes traveling to the city center, and arriving forty minutes later'. This may also do other pickups and drop-offs on the way but will arrive more or less when it says unless there are problems.
Some are a pain, most aren't.
A lot of titles have nice (often 3rd party) mod managers now which'll download mods, keep them updated, and allow you to selectively disable them- all from a simple GUI. When I recently replayed Skyrim I spent about an hour or so browsing through all of the mods on Nexus, and installing those that look interesting, and in my mind it made the game a lot more fun.
I admit there were a couple that were slightly broken in odds ways, such as a custom weapon which was heavily overpowered, but it didn't take much effort to just turn them off again.
If they're taking money for it and selling it openly to the public then it's been released, it doesn't matter that the technology is labelled as being 'for development'. Were they to restricting sale to people who qualified as a developer in some way then maybe, but as it is most of these are finding their way into the hands of people who use them to play HL2, another couple of games, and then take up some space in the wardrobe.
Dev companies won't care too much, they'll take the $120 hit on a device if it's something they're going to make a lot more from. Getting the paperwork sorted to buy them may be more hassle, but most will do it.
A lot of the mobile games are written using Unity which allows C# to be used on all platforms, from mobile, through PC, and even consoles if you have the dev license for them.
The only issue with Unity is if you need the Pro version be ready to pay some serious money, but for a lot of mobile games the Indie/Free version is fine.