The AI bot you need is one that has a fucking clue what to do re Brexit.
The future is AI, if Microsoft is to be believed. Microsoft UK CEO Cindy Rose wheeled out a succession of speakers on the topic at the company's Future Decoded shindig in London this week, with everyone from banking representatives to ocean researchers and the NHS on hand to extol the tech's virtues. A warning was issued to …
Yes, and if you recall the Cons solution to the Irish border involves (nonexistent) AI and sensor technology, hence the connection.
Johnson is the man who pointed at an EU funded fusion experiment, claimed it was British, and announced funding of less than 1% what would be needed to make fusion practical.
Artificial intelligence? We can't even get the right natural stuff to do important jobs, what hope have we got?
You can't predict it using logic because they people responsible for the decisions are not behaving logically.
A meaningless sophomorism. That's not how probability works. Aggregate outcomes do not necessarily have all the attributes of contributing factors. Indeed, usually they can't.
Also, it has yet to be demonstrated that human behavior isn't deterministic, regardless of whether it appears "logical" or "rational".
I agree. Last thing I need is my laptop figuring out how to use weapons and then noticing that I've started writing a letter and demanding that it get involved.
Clippy 9000: I see you're writing a letter, if you'd like me to remove this GLOCK-9 from your temple, follow these instructions.
Me: *sweating* Ok, ok...you win...let's go "Dear Sirs..."
Clippy: No, no..."To whom it may concern"...
Me: But Clippy...
Clippy: *pushes GLOCK harder*
Me: FUCK...ok, ok, ok....relax. "To whom it may concern"...
Kids: Daddy...why is the computer holding that to your head?
Me: It's ok kids...Daddy is just working from home today...just leave the room and find mummy, ok?
Clippy: ...and if you call the cops...*explosion sound*...right...*pushes on the GLOCK*...back to work..."to whom it may concern*...
Are both solutions looking for a problem. Not all companies have a problem that these feats of engineering are required to solve, either physically or economically.
AI can be badged on almost any rules based software decision making too. From the movement of ghosts in pac-man to making healthcare diagnostics solutions.
The statements made by M$ annoy me because they mean only "buzzwords get budgets approved" ....
Quite rightly said Giovani
Firstly the best definition of intelligence, is the ability to focus on different things at the same time... An average iq human can focus their concentration on about 3 or 4 things higher iq's 7 to 10... ML has the ability to focus on millions of individual inputs concurrently, and is at best an augment of human intelligence... Bearing in mind that human intelligence can't agree on what sentient life is.
Secondly the fact that M$ will keep banging on about AI... and marketing types and the more we'll heeled businesses will continue to run head long into new and shiny, is at best a worry. Like some how the crap that has come out of M$ in the last 10 years will all be OK because now it's artificially intelligent...this from a company that became such a behemoth from said marketing types that it no longer knows its brain from its backside... Nor can it release a working update for that matter.
Finally to what end does all this "INSIGHT" (shameless use of key buzz word) lead to with regard to the betterment of humanity?.. While there is no doubt about the potential for this technology it remains, for now anyway in the realm of humans. Humans who have lost the ability to choose intelligent leadership or intelligent use of hindsight.
Firstly the best definition of intelligence, is the ability to focus on different things at the same time
First, "first" is already an adverb (as well as an adjective, a noun, a pronoun, and if you really want to be difficult a verb). There's no need to suffix it with -ly.
Second, in what way is "the ability to focus on different things at the same time" a definition of "intelligence" at all, much less the best one? There are myriad mechanical systems which incorporate feedback loops for multiple factors.
An average iq human can focus their concentration on about 3 or 4 things higher iq's 7 to 10
Rubbish. The IQ metric is largely useless; what it primarily measures is an ability to do well on IQ tests. (I say this as an accomplished test-taker myself.) And while a correlation has been demonstrated between working memory capacity (WMC) and what cognitive scientists call "g factor" (for "general intelligence factor"), it's only one component, it's not clear how much of that is an artifact of the methodologies for determining g factor, and it's not clear what g factor actually means in practice.
More generally, treating "intelligence" as a single attribute that can be meaningfully measured by a single scalar value has been shown time and time again to be reductive to the point of uselessness.
All that said, what any of this has to do with the remainder of your post is unclear, since you then appear to go on to claim that ML can "focus on millions of individual inputs", but is not intelligent. So you've just contradicted your own claim.
It's true that not all companies have a problem for which AI/ML are suitable, though at a high level all companies are run by people making decisions which IS a role for these technologies.
But to conflate that with "it's a solution needing a problem" is clearly a terrible argument. Not all companies need super-computers, but developing ever better super-computers is not a solution looking for a problem.
Not to mention that science as a whole is OFTEN generating solutions to unknown problems. So many real life applications are built on stuff that was pure academia and turned out to be useful.
If MS Engineers say your 9's should be 0's then your 9's should be 0's and you should be thankful for it. They simply know better than you.
They patch and reboot your PC when they want.
They download your personal info and data from your PC whenever they want.
They monitor your PC usage, who you talk too, where you go on the Internet and what you buy.
They install software automatically without your say so.
They use your bandwidth as they see fit.
They turn on services you disabled, because you really didn't mean to do that.
We're obviously incapable of making adult decisions by ourselves. Thank you Microsoft. Thank you.
There's no guarantees with Access either when Microsoft move the goalposts. When Microsoft first introduced spell checking into Access databases it sneakily and without anyone noticing, changed all the single, lower case letter "i"s to uppercase "I"s. This was actually a data field and index to data in another database. The fault came to light after shipping out hundreds of software CDs to customers when the software fell over when certain records were accessed due to the broken index field.
Or really any application where strings containing digits and operators are not actual numbers. Or times can be mistaken for dates. Actually, despite using Excel a lot and finding it incredibly useful, I hate the stupid guesses it makes/forces on you when it comes to pasted/loaded data. It's just inappropriate if you need to get out exactly the data you put in, which is a shame, 'cause it could be great for a lot of ETL situations where sizes are not excessive.
"Actually, despite using Excel a lot and finding it incredibly useful, I hate the stupid guesses it makes/forces on you when it comes to pasted/loaded data."
This may make me sound like a Luddite, but if I need to do basic table-based manipulations (sorting, calculating totals) I tend to use Word (other word processors are available) exactly because it doesn't try and do clever things with your data.
I downvoted this and I don't believe in downvotes without explanations.
In point of fact, most spreadsheets make perfectly usable SMALL databases. Most are sortable and searchable. For many people, much of the time, that's all they need. Excel, which is prone to capriciously alter the user's data, is the exception, not the rule.
BTW, the much simpler Microsoft created spreadsheet distributed in early versions of MS Works worked fine for undemanding situations and did not surprise unwary users by helpfully changing their data in unpredictable ways. Personally, by the time I finally encountered a problem beyond the scope of the works spreadsheet, I'd had sufficient experience at work with Excel and its peculiarities, that I downloaded Open Office -- a choice I never regretted.
"In point of fact, most spreadsheets make perfectly usable SMALL databases"
Indeed. This is primarily what I personally use spreadsheets for (I'm not a bean-counter, so they aren't much use to me in their primary role).
I know how to use a real DBMS -- I use a variety of them daily in my work, and use them personally at home. But for a lot of things, a spreadsheet is just as good and a whole lot more convenient. The right tool for the right job, and all that.
Sure. For people who can't understand databases, and then find their data were somehow not what they thought the inserted.
I've seen this over and over... and spreadsheets used for tasks they shouldn't be used for is one of the big issues of IT.
And for the matter, any decent email client comes with an embedded contacts database.
Are A.N.Other El Reg Bots Boris AI Fans of his Zany Direction of Future Earth Space Travel ....... the Politically Printed Pictures that Media Machines Present to Tomorrow's Augmented Virtual Realities. And, take away the real news and what do you supply ..... apart from Just Premium Sub-Prime Bull Market Shit?
Carpe Diem, Boris. Oh, and you are not alone, for there are surely more than just a few enamoured of the refreshing style.
It was generally not an inspiring day, lifted by the closing hour. The opening keynotes were interminably dull with some frankly dubious sounding stats on AI adoption. The sessions I attended were much less techie than I remembered from previous years. Perhaps I just chose badly. However, Sharman and Peake were good. I could happily have looked at their photos from space all day, although I also enjoyed what they said. Sharman in particular.
Due to the complete lack of intelligence of some people that I have to work with.
If I knew how to write a bot that could fill in a simple form several times a day it would make my life a lot easier than the pesky humans that log every call under 'genetic fault' rather than an error code and can't take down a contact number without buggering it up. Even customers that have access to the system can manage it some of the time.
I would settle for HAL 9000 at this point but probably not give it access to the entry system for obvious reasons.
It may not be the people that you work with that are the problem. We had an office that was only manned for part of the day and people were encouraged to leave a message at other times. I would listen to the various messages that rambled on about their requirements, their current lifestyle and sometimes that of their pets but the really important information, the contact number, was tagged on the end in the manner of an inarticulate racecourse announcer in the final furlong.
"but the really important information, the contact number, was tagged on the end"
I don't know why this surprised me, but it did.
When I'm leaving a business message, I state my name, contact info, and (if appropriate) account and/or invoice # at both the start and end of the message. I guess this is just one of those things that seems so obviously a good idea that I thought most people do this. That's an interesting blind spot, as when I think about it, it also seems obvious that a lot of people wouldn't. People gotta people.
This post has been deleted by its author
I do the same. Contact details first. Anybody that's had to listen to a message of several minutes over and over because they can't quite make out the number at the end soon realises that putting the number first so if the listener can't quite make out what the number is they don't have to listen to the entire message before they get to the number.
I wrote a program that checks pdf documents then emails them back to the person that submitted the document with a list of their errors.
One surprising thing about this method is that humans appear to prefer to be told they have made a mistake by a robot rather than another human. Using a robot to train humans is a amazingly effective in reducing the errors humans make.
What borked dataset will be use to identify any aliens we meet? I expect the usual one over filled with caucasians won't cut the mustard in that scenario. Telling the aliens they all look the same to us probably won't go down any better than when we used it on Indians or Chinese. Though my wife works with a guy she says is a dead ringer for Pooh Bear.
Oh dear I hope El Reg's distribution in the People's Republic won't be affected.
>>"We need to start building up the trust in AI"
Maybe if we could see AI applied to a problem that directly impacts us?
For example, I would be interested in an AI system that would improve Traffic Control.
The only thing I really know about current traffic control systems in my neighborhood is that they are inefficient.
AI I don't know so much, but I believe traffic lights have been programmed with fuzzy logic for some time.
And it does work, except for the idiots who have to move into the middle of the junction so the turning lane sensor doesn't detect them then turn against a light which is now red.
" "to fully capitalise on its standing as a leader in AI"."
What so your company can spout out bullshit marketing AI and Machine Learning bollocks?
FFS. Getting really annoying now. We were on a webinar call today for a CCTV system based in the US of A. Looks nice but the amount of times AI and Machine Learning was mentioning was a little annoying. But they didn't mention it too much. Ironically at the end with the Q&A someone asked if the cameras could do ANPR reading. Skip the part where we had to explain to the American's what ANPR was (they have a different name for it) but the answer was funny. "Yeah we still are looking into this. As its not a huge market..." What? Its a massive market, we didn't say anything though. He continued "..this would mainly be a software thing and we're not quite there yet" what? But you've just spouted AI and machine learning. Surely if you've implemented such then ANPR recognition would be a piece of piss. Or our you just using those words as marketing bullshit. And why, when doing presentations do US salemen talks to fucking fast. Stop and take a fucking breath. Its not a fucking race. Is it to hide your bullshit in noise? In the hope we didn't hear because you said it so quickly?
Thank you for your query.
Our customers are very important to us and we hope we can help you with using MS Powerpoint™
To better understand your situation please tell us which Windows edition and version you are using and similarly for your Office installation.
Have you recently installed any other software or made any significant changes to your system?
Have you checked that your Locale is set to Low Earth Orbit and the Date and Time are correct?
Senior MS Advisor Partner
It's not there to store data or numbers or anything like that, that should be obvious to anyone using it.
It would be nice if Microsoft would release any indication on what Excel is supposed to be. If I was allowed to put on my tinfoil hat I'd say they won't, because that would mean they would have to actually respond to bug reports.
Before I trust that MicroSquish has any grasp of artificial intelligence, I need proof that they have some grasp of natural intelligence. That their business applications, product, and marketing groups could be replaced -- to universal benefit -- with a detached switch in the off position demonstrates all the proof I need.
Speaking of which, I think I have something here with about 94 proof that will suffice as defense against all the torture MS can render on this fine day. Tomorrow will have it's own MS terrors. Cheers.
It's an odd statement without more context. A bit like saying "companies that give employees free biscuits are outperforming those that don't."
I bet companies which give free biscuits DO outperform
-care about their staff more which is probably reciprocated
-care about small details
-willing to spend money on things beancounters may say are unnecessary
Biting the hand that feeds IT © 1998–2019