Microsoft bungs a billion bucks at biz developing AI that will take our jobs 'for the benefit of all'

Microsoft has plowed $1bn into OpenAI, a San-Francisco-company that is working on artificial general intelligence (AGI). OpenAI is a nonprofit-for-profit organisation founded in October 2015 by investors including Elon Musk and Sam Altman, the latter now CEO of OpenAI. According to a statement today, the two companies have …

[Steve Davies 3]

For the Benefit of all???? ROFL

Sorry SatNad, the only one to benfit is Microsoft. Everything you do is for the benefit on MS.

But I'd like to know what his answer to the following question is.

Hey SatNad, when all the AI Systems and Robots have taken our jobs who will be able to afford to feed themselves let alone pay for MS products and services?

Come on now. Don't be shy! Tell us what you think...

[Anonymous Coward]

Re: For the Benefit of all???? ROFL

Dear Messrs. Hargreaves, Lea et al, when all the Spinning Jennies, Stocking Frames and the like have taken our jobs, whoso shall afford to feed themselves let alone purchase thy products?

I am Sirs Yr obliged & obedt Servant (not),

N. Ludd

[ThatOne]

Re: For the Benefit of all???? ROFL

> who will be able to afford to feed themselves let alone pay for MS products and services?

Easy: The few wealthy companies who will be delighted to run (almost) without annoying staff, and who sell to other wealthy companies also running without annoying staff.

It's not like Azure is targeted at the Joe Sixpack in the streets. The (wo)man in the streets is only good to deliver them pizzas and wash their car, and they're working on automating that too.

[Halfmad]

Re: For the Benefit of all???? ROFL

It's almost as if they are a private company and not a charity.

[Anonymous Coward]

Obviously...

"Cortana, open the pod bay doors"

[Halfmad]

Re: Obviously...

"Shuffling songs by the Doors".

[Starace]

Good luck

At what point do they discover what's actually hiding behind the usual Musk smoke and mirrors?

Still waiting to be convinced they have anything new, interesting or innovative compared to the rest of the market. At the moment they just seem to have a spin on the usual old stuff and even that feels thin once you scrape off the hype; GPT-2 certainly didn't seem like anything special or even particularly capable.

[DCFusor]

Re: Good luck

"GPT-2 certainly didn't seem like anything special or even particularly capable."

Sadly, much the same can be truthfully said about quite a few humans, who nevertheless can become quite rowdy when hungry.

Think of all the journos - for just one - who do nothing but copy-paste AP and perhaps change one word?

(Not present company of course)

[I.Geller]

Parsing the clause "Fekla and Ruclan are singing" OpenAI gets only one pattern "Fekla and Ruclan are singing"

Meanwhile my patented AI-parsing gets from the same clause three patterns including all whatever OpenAI gets:

- Fekla is singing

- Ruslan is singing

- Fekla and Ruslan are singing

This is caused by the use of 75-years old n-gram and not of my patented AI-parsing.

[Anonymous Coward]

?Who's gonna stop the fake news when the polls are down?

"could be used to generate news reports that were hard to distinguish from human-authored ones, a dangerous tool in an era of fake news"

Or a very useful tool, if the polls are down on the candidate/political party that the boffins support. No need for shadowbanning, just bring on the fakery!

[JassMan]

Re: ?Who's gonna stop the fake news when the polls are down?

"could be used to generate news reports that were hard to distinguish from human-authored ones, a dangerous tool in an era of fake news"

What the world needs is AI with a conscience. Unlike current AI which is as biassed as its training data, if the starting point in training was a conscience, a good AI would look at fake news and hunt down all the little falsehoods which make fake news look so plausible on the surface. Something which most modern journalists fail to do when re-reporting untruths and giving just them that bit more legitimacy.

We wouldn't be in the Brexit mess we have today if journalists hadn't republished Boris's story about bent bananas, instead of actually reading the directive and commenting on the truth (to give but one example). The problem with fake news is not that most intelligent people can see the lie, but that most of the general public see the grains of truth, add them together and make a cake to keep as well as eat.

[Anonymous Coward]

Re: ?Who's gonna stop the fake news when the polls are down?

Brexit is opposed to centralisation of power, so is consistent with the objectives of a consciencious AI.

There's always been mainstream fake news, but today there's more alternative fake news, plus alternative true news that criticises mainstream fake news, plus mainstream double-fake news that selectively reports on fake alternative news that was created soley by mainstream actors for the purpose of undermining all alternative news.

[I.Geller]

Re: ?Who's gonna stop the fake news when the polls are down?

I have some hints of conscience and I am at least de jure right now and, with some help, can de facto own AI and control its future... Sergey Brin and others... You decide...

[amanfromMars 1]

Re: ?Who's gonna stop the fake news when the polls are down?

I have some hints of conscience and I am at least de jure right now and, with some help, can de facto own AI and control its future... Sergey Brin and others... You decide... ... I.Geller

Oh please, I.Geller, surely you know that is always to be something you decide, but/and you will need to be significantly better than just good and great at owning/pwning/preening AI for control of the future, for competition and opposition will ensure it be so.

And as for ....

The OpenAI charter ..... came under question in February this year when OpenAI developed an AI model, called GPT-2, which it was claimed could be used to generate news reports that were hard to distinguish from human-authored ones, a dangerous tool in an era of fake news.

"Due to our concerns about malicious applications of the technology, we are not releasing the trained model," the company said.

Good luck with trying to keep that sort of development out of the mainstream. The battle has surely long ago been lost. Just read daily newspaper headlines and the tales they introduce to realise the truth and everything new and revolutionary is feared by owners of the past who have abused and misused and taken full wanton advantage of human ignorance and how easy the masses are led by media devices giving suggestive audio-visual signals.

The fundamental error that has been made by that and those in present positions of extreme power and past influence, is that they never realised nor imagined that others around them would be a great deal more intelligent and more knowledgeable than they would ever be enabled to be, and that they would discover the secret ways and virtual means to have memes come alive and rule in a world full of worlds hosting minnows with wannabe sharks.

And all that is available for either attack or defence are the uncertain Tell Tale Terrors of FUD.

[Teiwaz]

I suppose we should be grateful

The 'old' microsoft would have just bought them (then proceeded to dinker around with it until what made it valuable was flushed)..

This way, an open business gets some needed cash, and MS gets another B-J.

[Inventor of the Marmite Laser]

But

Will they actually TEST any of the software.

At all

[Anonymous Coward]

Re: But

> Will they actually TEST any of the software.

No.

[I.Geller]

Re: But

I tested it long ago: AI can be much worse you dare to imagine.

[Anonymous Coward]

Somehow our future seems to be developing into a scene from Dune.

[N2]

No worries

Judging by Microsfts commitment records to anything useful.

[macjules]

Does your AI bite?

Oversight comm: "OpenAI, does your AI really work?"

OpenAI: "Yes, of course it does"

Oversight comm: "But it has killed all those Tesla owners and pedestrians"

OpenAI: "But that was not our AI. We work for Microsoft now."

[JulieM]

Not releasing GPT2

FTA: "Due to our concerns about malicious applications of the technology, we are not releasing the trained model," the company said.

As though that would stop anyone from just re-implementating the whole thing from scratch. Their programmers aren't any smarter than a million independent experimenters, who also have access to all the same training data.

[Starace]

Re: Not releasing GPT2

More because if they really let it out in the wild it would be even more obvious that it spouts even more gibberish than a Buzzfeed reporter - it's only convincing if you don't read very far and are really bad at spotting the inconsistencies.

There are better systems out there.

[Anonymous Coward]

Latest windows 10 update required two re-boots, FFS MS spend it on your shit OS first and not making updates a torture.

[mr-slappy]

Plowed?

"Ploughed," please.

"Plowed" is "N. American or archaic" (possibly the same thing?) according to the Chambers Dictionary on my phone.

[I.Geller]

OpwnAI AI does't understand us, where

understand

verb

1.

perceive the intended meaning of (words, a language, or a speaker).

"he didn't understand a word I said"

2.

interpret or view (something) in a particular way.

"as the term is usually understood, legislation refers to regulations and directives"

My AI understands us because it compares patterns, which AI gets parsing our dictionary definitions.