That why you train your algorithm on all the available cases but the last year ones. Then you can test it on that last year of cases. For the system the last year is the "future" on which you do your testing.
When the european commission does a public consultation, it get aound 300 comments (even when the policy could have worldwide implications). Most of them are from companies trying to keep some kinf of status quo, then from NGO saying it doesn't do enough and very few people answer under their own name (eg: DG CLIMA consultations. In this case, 500 comments is a very good return rate...
I have written an expert system for an oil company for predicting output of distillation towers at 1, 2, 4 and 8h in the future. The V1 of the expert system has the rules in a database and is working perfectly but is a bit slow. For the V2 of the system we auto-generated some code from the rules to include them at compile time. Do you say that V1 was a real expert system but V2 is just a fancy prediction software? For me they are the same. They still use the first version to add/debug rules before compiling a new revision of the V2...
I have seen an article saying that the current LiDAR they use cost around 70.000$ and they hope that the mass production of vehicles will reduce its price.
Asynchronous designs are faster (~3x) and consume less energy (~2x) but need an overhaul of the production process who is deemed too costly. Perhaps this technology could make it interesting again. (Source)
And it is that mentality that make working as software engineer such hassle. These 10% are as important than the other 90% event if they take 90% of your time on the whole project. It's like a house. You have done 90% of the work once you have the wall, all the electricity and the plumbing ready. Sorry but these plaster panels and paint job are just for the perfectionists, they are not necessary to live in your house! But you still expect them no? This is exactly the same for software. You expect your software to work for every use-case because to implement that last use-case you need to rewrite or restructure a part of the already existing software. You expect it to work on the million records database and not the meager 20 records used for tests without needing a data center to run your program. I have seen too much of these corner cutting things from manager, saying that I am a perfectionist, who made what should have been a nice project into a living hell.
You can find good hand made mechanical watches from 1000$ but they are still not what I could call high-end. At that price you start to expect to not find someone else with the same watch.
I have a mechanical watches because:
1) They are damns fine piece of mechanical art and craftsmanship
2) Seeing them ticking is mesmerizing
3) I am pretty sure that I will never encounter someone else with the same ones
I will never buy a "smart" whatever crap that every one else has. I want my watch to show who am i. I don't want a watch that say: "I am like everyone else". And for the "smart", I already have a smartphone
What a piece of plastic and silicon build by the millions in a fab will say about you?
Yes there is resistance, that is human nature to first resist to new ideas. And you show with your 2 first examples that it can take time for the community to accept evidences.
Cold fusion is still looked after and the last research demonstrate that there is no need to change the nuclear theory. There are new theories and hypothesis coming and going (like the Widom-Larsen hypothesis). But it is still confidential and the scientific publications are not eager to publish on that subject after the the 1989 polemic.
In 2011 the OPERA experiment reported consistent and repeated observations of faster-than-light neutrinos and again the physics community questioned the observations instead of modifying or rejecting special relativity. And, like cold fusion, it was wise of the physics community to question the OPERA results.
It was wise to question the results because later they found that there was a problem with a connector of a high precision clock and once that connector was changed everything was back to 'normal'. They didn't 'corrected' the data, just replaced a faulty connector but there is still some conspiracy theorist that will tell another story
I would add that for ESP and other psychic phenomena, James Randi still have 1 million dollar for you.
To kill hft, put a 0.001% tax on every transaction especially the cancelled ones. And hft will stop over the next nano second.
Why Google didn't reacted following the Apple case? It was just a question of time before the same kind of lawsuit would begin against them...
Officially, war never stopped between the 2 Koreas
Another point of the story. Don't reuse passwords
But money will not be the same. You can imagine a society where everyone receive enough money every month to be able to live and enjoy it. You can have more money from the 'state'/'governing body' by doing necessary thing that the machines can't (yet) do (Farmer, Teacher, Doctor, Police...). You can have more from your fellow humans by providing some kind of service (Cook, Artist, Performer...). You can also create new things/products that machines can produce and receiving some money for each asked by someone and produced by the machines.
The 'state'/'governing body' will control the machines.
The lazy would be lazy and live a decent life. The productive would be productive and live a better if not the best life.
No, they are good enough even for gaming. It is true that i can't activate every little graphic options and set them to the max. But my 3 year old laptop is good enough to be able to play any recent game.
Now if you are one of these player that think that 160fps is not enough and they need mooooaaar power even if their screen can display only 50-60 fps. In that case, even your desktop should not be good enough.
You just don't need to remember 1 password, but 11 of them to log in... What an improvement !!!