|
It is true, for real scientific theories. It isn't for "belief systems" dressed up as theories.
For theories like Evolution, General Relativity: they make predictions which can be tested. Events happen, which can be checked against the theory and either confirm it, or disprove it. Quantum mechanics will probably disprove some of GR, when we learn how to do proper tests - but it works for the moment on a large scale.
For a belief system, anything which contradicts is shouted down, or ignored: Intelligent Design, anybody?
Regrettably, as the mind becomes older and more rigid, a Theory tends to slide into a Belief.
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
AI is such a bloated topic. A few years ago, IoT and cloud and "connected" objects were the new black, now it is AI - and frankly, 99,99% of things that are called "AI" are not even close to anything about artificial intelligence at all... Even a PID controller is artificial intelligence by today's standards. It is crazy how this damn freaking "ultraconnected" "social media" "buzz-addicted" society is going to ruin everything...
So honestly, no wonder Cloud in AI has no success.
|
|
|
|
|
Me was wondering always how the 'I' part is actually got it's name... We not even know how Human Intelligence works (not even agreed on what it is), but already building the Artificial version?
"The only place where Success comes before Work is in the dictionary." Vidal Sassoon, 1928 - 2012
|
|
|
|
|
Kornfeld Eliyahu Peter wrote: ... but already building the Artificial version? Starting right with the Holy Grail? Has this ever at any time or place led anywhere? At least we will probably not have the Terminator knocking at the door anytime soon.
Kornfeld Eliyahu Peter wrote: We not even know how Human Intelligence works (not even agreed on what it is) No wonder. You have more neurons in your head than there are stars in the galaxy. These neurons already start processing inputs and forming their network while you are still being assembled. Since you don't get a fresh installation of Monkey 10 (after Monkey 8 was a failure and Monkey 9 was skipped and many still hold on to Monkey 7), it looks like this thing actually programs itself. You need a little more than just a heap of neurons to do that in a fairly deterministic way. The real thing is the product of many million years of evolution. Artificial evolution of the 'hardware' may need time on a similar scale.
I have lived with several Zen masters - all of them were cats.
His last invention was an evil Lasagna. It didn't kill anyone, and it actually tasted pretty good.
|
|
|
|
|
Not only are we (i.e. humans corporately, not CPers) aiming for Artificial Intelligence, but we are looking for extra-terrestrial intelligence (SETI). The real reason must be that we have given up looking for terrestrial intelligence of any kind (excluding dolphins).
|
|
|
|
|
I would not want a computer to operate in the same manner that 'intelligence' operates.
You know those "optical illusions"? Those are your brain processing visual inputs quickly to come to the wrong conclusion so that you can avoid that tiger stalking you in the jungle.
Neil deGrasse Tyson calls "optical illusions" "brain failures".
I wouldn't want a computer system to be as poor at processing data accurately as my brain is.
“That which can be asserted without evidence, can be dismissed without evidence.”
― Christopher Hitchens
|
|
|
|
|
GuyThiebaut wrote: I wouldn't want a computer system to be as poor at processing data accurately as my brain is. Then better don't look up what the word 'heuristic' actually means.
I have lived with several Zen masters - all of them were cats.
His last invention was an evil Lasagna. It didn't kill anyone, and it actually tasted pretty good.
|
|
|
|
|
I know, which then brings up the area of non-deterministic functions/methods/algorithms.
So for the same input you could be getting different results, or for a very small difference in input you could get a very large difference in the result.
Computer systems have tended to be seen as very much deterministic systems and with the advent of AI we hit the problem where it may not be possible to explain or justify why a result was reached because of the underlying complexity of weightings.
Which means that we would need to develop tools that could analyse and explain why a certain result had been arrived at, which would also incur a massive overhead of saving all versions of weightings/algorithms for each result. As if it was a heuristic system that changed its weightings/algorithms at will the original 'checked in' code may not have been what was used to produce this current result.
“That which can be asserted without evidence, can be dismissed without evidence.”
― Christopher Hitchens
|
|
|
|
|
Kornfeld Eliyahu Peter wrote: Do we don't like the 'cloud' idea? I don't like it at all. If I want to play mainframe, I expect to get a 300 baud VT100.
Kornfeld Eliyahu Peter wrote: Do we prefer other services than AWS used in the tutorial? Yes, indeed. For AI I prefer the services of my GeForce chipset on the graphics card.
Kornfeld Eliyahu Peter wrote: Or we just simply to bricked an old to (re)learn new/old tricks? If something does not interest me, it does not interest me. That has nothing to do with age. I already was like that when I was 12 and it has not changed. The list of things that once were 'in' and have died without ever getting any attention from me is long. BASIC, anyone?
I have lived with several Zen masters - all of them were cats.
His last invention was an evil Lasagna. It didn't kill anyone, and it actually tasted pretty good.
|
|
|
|
|
CodeWraith wrote: That has nothing to do with age
Probably - we are old enough to know that things of fashion tend to die faster than the autumn fly...
"The only place where Success comes before Work is in the dictionary." Vidal Sassoon, 1928 - 2012
|
|
|
|
|
Even if they don't, they still don't interest me and genuine interest is the fuel for anything where I ever was sucessful.
I have lived with several Zen masters - all of them were cats.
His last invention was an evil Lasagna. It didn't kill anyone, and it actually tasted pretty good.
|
|
|
|
|
CodeWraith wrote: BASIC, anyone?
Now that is not a good example.
Written from my TO7
|
|
|
|
|
Interpreters like BASIC were almost the only languages that were practical on 8 bit and early 16 bit computers. I saw them as a dead end and it took me another ten years before I had a computer that had the memory and mass storage resources for a C compiler. What has remained is my enduring dislike of interpreters of any kind.
I have lived with several Zen masters - all of them were cats.
His last invention was an evil Lasagna. It didn't kill anyone, and it actually tasted pretty good.
|
|
|
|
|
There is a huge financial risk in signing up to cloud services at present.
There are a number of financial analysts who have tried to dig into Microsoft's Azure numbers, and it's not easy as Azure gets lumped in with a bunch of other profitable stuff.
The risk is that the cloud providers are currently attempting to buy market share with unsustainable prices to grow a client base. At some point, profits will be required and prices will rise.
There are also increasing technology lock-ins being created in the cloud offerings, which
make it harder to transition from one provider to another. The cloud has moved from being a remote virtual server offering to being a technology stack offering, and these stacks are not compatible across cloud vendors.
|
|
|
|
|
|
I'm very interested in the cloud, mostly Azure.
Not very into AI though (which doesn't even exist, it's just a marketing term for ML).
And I try to stay away from SAP.
I wanted to participate in the serverless challenge a few months back and I even started writing, but something came up and I couldn't find the time after that
The result of that challenge was somewhat disappointing, as all entries ultimately discussed Functions, while other serverless services such as Service Bus, Logic Apps, serverless API Gateway and the SQL Server serverless offering were completely ignored.
Every time I see serverless mentioned with Functions or Lambda I think "not again... "
Two months is a long time, but writing an article is no small task either
|
|
|
|
|
Nice... What actually you find interesting (more interesting?) in cloud than in the more traditional server setup?
"The only place where Success comes before Work is in the dictionary." Vidal Sassoon, 1928 - 2012
|
|
|
|
|
That I don't have to do system administration (which I can't and providing I stay PaaS/SaaS).
This is especially useful for some of my smaller customers who don't have a system administrator or even a server.
I build the software and deploy it and it runs for a small fee.
Spinning up Azure SQL is SO much easier than spinning up a server, installing SQL Server, securing it... Same for web apps and IIS and pretty much everything else in the cloud.
I can deploy an entire environment, including database, web apps, queues, everything, fully automated using nothing but some relatively simple scripts.
Scaling in the cloud is easy and mostly automated, especially when you're using serverless.
Serverless wasn't even a possibility before the cloud.
In a recent project I was able to set up everything in Azure in a matter of days.
A small sync with the on-premises server took days because I needed to wait for the server admins (who messed up a couple of times).
Especially PaaS is easy as you have no infrastructure at all.
No updates to worry about either.
Not to mention the services is offers.
You need a queue? Here's Service Bus.
Want a web app? Here's App Services (with Application Insights, slots, authentication...).
Need some regular jobs or event triggers? Go for Functions and/or Logic Apps. They easily integrate with many other services as a bonus.
Going for a little API management? Here's API Management.
Need a database? Why not try the PaaS Azure SQL, PostgreSQL or MariaDB? Or Maybe the NoSQL CosmosDB is more to your liking?
You need to store those connection string? Azure Key Vault is the solution...
Logging, monitoring, backups, redundancy, mirroring? We got your back with various services and options!
And when all else fails they still offer various VMs with an image to your liking.
Oh yeah, it all integrates with Visual Studio and your .NET Core applications so don't worry about that.
People who warn about security issues ignore the fact that who knows how many on-premises networks get hacked too.
In fact, I bet the cloud is more secure, when you deploy it correctly, since you have computing power that an on-premises networks will never achieve.
There are actual teams that try to hack and secure these cloud environments.
After a few weeks these teams switch places to see if the previous security programmers can hack their own security.
How many companies do this!?
Should you really worry about security you can always integrate with your on-premises network.
And yes, if you use admin/admin as a username/password combination[^] you're just asking for it, no matter where you host
People who don't like the cloud just don't know the possibilities
I can think of very little reasons not to move to the cloud.
I currently have one, which is Crystal Reports, which doesn't run in App Services and would require a VM.
However, that's more like a reason to move away from Crystal Reports than from the cloud
|
|
|
|
|
I did one of the challenges a while back. I also did the recent AI one but early on it wouldn't work and I don't have the time to try and troubleshoot and learn new tools.
Kornfeld Eliyahu Peter wrote: which maybe because it gives you a $50 prize for almost nothing I didn't see anywhere that mentions this. But my time is also worth more than $50.
However, I am incredibly impressed by the amount of effort that goes into creating these challenges.
Social Media - A platform that makes it easier for the crazies to find each other.
Everyone is born right handed. Only the strongest overcome it.
Fight for left-handed rights and hand equality.
|
|
|
|
|
ZurdoDev wrote: incredibly impressed by the amount of effort that goes into creating these challenges
Definitely. That's one of the reasons I'm interested in why this one (seems to be even more planned the the previous ones) is mostly ignored...
"The only place where Success comes before Work is in the dictionary." Vidal Sassoon, 1928 - 2012
|
|
|
|
|
Kornfeld Eliyahu Peter wrote: mostly ignored... I can only answer for me, and I had not even noticed it. It looks like an ad on the front page so I ignore it. I didn't know about it until I read your post.
Social Media - A platform that makes it easier for the crazies to find each other.
Everyone is born right handed. Only the strongest overcome it.
Fight for left-handed rights and hand equality.
|
|
|
|
|
Kornfeld Eliyahu Peter wrote: Why is that so?
It looked interesting but:
1. Time constraints were a factor
2. My own project (not AI related) is more interesting
3. While there are some really cool things, like self-driving cars, I simply don't buy into the PhD/PhB approach to AI. But what else is new, right?
|
|
|
|
|
Dividing line removes damage to particle? (11)
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
Removes damage = de-mar (slightly dubious)
Particle + cation
DEMARCATION = dividing line
Whenever you find yourself on the side of the majority, it is time to pause and reflect. - Mark Twain
|
|
|
|
|
Quote: slightly dubious Hence the question mark!
You are up tomorrow!
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
AntiTwitter: @DalekDave is now a follower!
|
|
|
|