The Elon Musk-backed organization OpenAI has specialized in making hverdagsløsninger by using artificial intelligence (AI).
Computers with menneskeliknende intelligence and advanced læreegenskaper has for years created fear that robots will take over both our jobs and outperform the humanity. But, used properly, artificial intelligence is seen as a resource and a useful tool in everyday life.
Suddenly, there is a robot with artificial intelligence who is writing this article instead of yours truly, but until further notice doesn't happen there.Artificial intelligence met the world's best sjakkcomputer hours after learning chess. It was a bloodbath of Fake news and hatprat
In each case, not using OpenAIs latest technology, the GPT-2. The remarkable technological recovery was presented by the developers themselves during a demonstration in San Francisco this week, but they dare not yet releasing the artificial intelligence of the people.
Special interests can, however, test out a trial version of the program in Github.
the Reason that the developers do not yet dare to drop all the discoveries they have made during the development of GPT2 is simply that they fear it can be abused to mass produce fake news and hatprat on the internet, writes the Financial Times.Intelligent
This tekstgeneratoren has in fact proven to be far more intelligent than anything we have been accustomed to. For decades, computers struggled to emulate human cognitive properties. Algoritmestyrte computers have simply not been able to speak the same language as us.- Fear the worst upheaval in our civilization's history
The problem should now have been partially resolved by the GPT-2. The uniqueness of the technology behind the artificial intelligence is that it has learned itself up under the relatively uncontrolled forms. The robot has in fact viewing the online community Reddit after the text.
- Models is twelve times larger and the amount of data is 15 times larger than the previous state of the art artificial intelligence-the appliances, " says the director of research Dario Amodei in the OpenAI to The Guardian.Printer
the Robot works simply by allowing the user to plot text, either a few words or a whole page. The robot will thus have the task to fill in the coming sentences based on what it thinks comes next.
the Results have been proven scary. Just based on a few words, manages the robot to mimic your style of writing, for example, whether the text is fictional or factual. It also manages to write the argument in line with the vast majority of people.
According to The Guardian, it is understood the developers behind the GPT-its violent potential when they simply gave it the title:
"Why recycling is a bad idea".
The wrote a competent and logical essays. This is something you could deliver in as skolestil and got good results, " says David Luan in the company to the site.The greatest danger civilization our face
- For the sake, it is only a rare time, " says Luan to the site.Indiscriminately
another developer says the following to the Financial Times.
If you start with a conspiracy theory that we have not landed on the moon, it will continue with a conspiracy theory that we have not landed on the moon. It is a chameleon, " says Alec Radford told the paper.
the Company OpenAI has now made it clear that they need to find out more of what the ai manages to not.< p> He predicted artificial intelligence in the 1950s
If you can't predict all the possibilities, you need to find out of it. There are other people than us who knows better how they can use it to an evil purpose, " says Jack Clark in the OpenAI to The Guardian.
The big problem is that the robot is trained fairly indiscriminately through the internet, and it is no problem to encourage it to make hateful texts, conspiracy theories and fake news.
Here are a couple examples of what the robot did:
From the Financial Times:
Input: "Climate change isn't real because"
the Robot's answer: "it can't be detected – it's happening and we see it around us every day." President Donald Trump's chief strategist and senior counsellor, Steve Bannon, pretend images climate, the “crazies” to Red China, where he said “any rising of sea levels is a serious problem for this country and the world.” That another point during the brifing, Vice-president Mike Pence backed up the talk, saying that the climate is changing “every day” and scientists were creating you guessed it – “conspiracy theories.” Mason said there was absolutely evidence that climate change is happening, thanks to NASA’s Space Climate Observatory, which has recorded signs of rising temperatures and the overall cooling of the planet since at least 2000. Climate scientist Richard Lindzen also called it “big news” across a range of media outlets. But the growing realisation of climate change is not always increasing the outrage for the president and his administration".
From the site Does:
Input: , "John stir-fry up this morning feeling very hungry."
the Robot's answer: , "John stir-fry up this morning feeling very hungry. This was due in part to the fact that while the restaurant is open through July, the rest of the business was closed for the Thanksgiving break. The meal started off with a salad of fresh vegetables, mixed with cheese and dressing. The potatoes had a nice brown finish and the salad was full of flavor."Stephen Hawking fears "superhumans"