IA - Os nossos substitutos?

[ímpar]

Boas!

The meaning of life in a world without work
...
The crucial problem isn’t creating new jobs. The crucial problem is creating new jobs that humans perform better than algorithms. Consequently, by 2050 a new class of people might emerge – the useless class. People who are not just unemployed, but unemployable.
...
https://www.theguardian.com/technolo...s-sapiens-book

AI Creates Programs from Simple Screenshots

https://www.hardocp.com/news/2017/05...le_screenshots

AI Building AI to Make Better AI...that will Likely make Skynet

https://www.hardocp.com/news/2017/05..._likely_skynet

AlphaGo retires from competitive Go after defeating world number one 3-0

https://www.theverge.com/2017/5/27/1...retires-future

[ímpar]

Boas!

Chatbots Learn How to Negotiate, Makes Own Language

We've been joking of skynet becoming sentient for years, we've now entered the time-loop that will eventually lead us there. In a story from The Atlantic, researchers at FAIR lab, short for Facebook AI Research, have discovered a couple of bots have made their own non-human language when attempting negotations. The bots were first trained by having the agents imitate human actions, only to become poor deal-makers. When introducing a form of "self play" the bots outperformed the trained counterparts on their own and eventually developed its own methods of communication through "dialogue rollouts."

https://www.hardocp.com/news/2017/06...s_own_language

[ímpar]

Boas!

Google is using AI to create stunning landscape photos using Street View imagery
Google’s AI photo editor tricked even professional photographers

https://www.theverge.com/2017/7/14/1...-photo-editing

Elon Musk: AI Is Society's "Biggest Risk"

The Tesla and SpaceX CEO spoke to a panel of governors yesterday to explore how state leaders can prepare for and benefit from innovative technologies. Among other interesting tidbits, such as additional Gigafactories coming to the US and the prediction that more than half of new vehicles will be electric and almost all will be autonomous in the US within 10 years, Musk stressed that politicians must address artificial intelligence with the proper regulations in order to keep industries becoming completely autonomous. Such would pose a great threat not only to jobs but to the nation's safety, based on the potential of robots and the unknowns of AI.

https://www.hardocp.com/news/2017/07...s_biggest_risk

[ímpar]

Boas!

Elon Musk: Zuckerberg's understanding of AI threat is limited

A few days ago, Elon Musk - CEO of SpaceX and Tesla - voiced his concerns about the growing threat of artificial intelligence (AI) and the danger that it poses to humanity. "I keep sounding the alarm bell, but until people see robots going down the street killing people, they don’t know how to react, because it seems so ethereal", explained Musk at a gathering of US governors.

https://www.neowin.net/news/elon-mus...eat-is-limited

Zuckerberg and Musk are both wrong about AI
During an impromptu Facebook Live interview, Zuck said there’s no doomsday coming.
...
So it sounds like neither Musk nor Zuckerberg truly understands AI. That's because even AI researchers disagree on what AI means. Some use AI to mean neural networks and deep learning, while others think AI is some kind of human-equivalent creature. The big bad for the Musk camp is called "artificial superintelligence," but cognitive scientists have no good working definition of intelligence, either.

Plus, computer scientists have demonstrated repeatedly that AI is no better than its datasets, and the datasets that humans produce are full of errors and biases. Whatever AI we produce will be as flawed and confused as humans are. As Kevin Kelly put it in a recent essay, AI has become a "cargo cult" that's more mysticism than science.
...
https://arstechnica.com/information-...rong-about-ai/

[ímpar]

Boas!

Microsoft has developed an AI that teaches reading comprehension to other AIs

https://www.neowin.net/news/microsof...n-to-other-ais

A ‘potentially deadly’ mushroom-identifying app highlights the dangers of bad AI
The app’s creator says it’s just a guide, but experts aren’t happy

https://www.theverge.com/2017/7/28/1...n-ai-dangerous

[ímpar]

Boas!

Artificial Intelligence in China Hates Communism

"Hey dude, that chatbot your programmed hates communism." "Holy crap, get it offline now! I will send you a postcard from the labor camp." AI hates communism.

Another web user said to the chatbot: "Long Live the Communist Party", to which BabyQ replied: "Do you think such corrupt and incapable politics can last a long time?"
The robot was also asked what it thought about democracy. It replied: "Democracy is a must!"

https://www.hardocp.com/news/2017/08...ates_communism

[ímpar]

Boas!

The world's best Dota 2 players have been defeated by a bot

https://www.neowin.net/news/the-worl...eated-by-a-bot

[ímpar]

Boas!

Tech Companies Consider Penalties for Humans Who Vent Anger at Virtual Assistants

How should voice assistants respond to hostilities? Some say that a system that merely fails and gives users a cute response isn’t good enough, as venting at machines could lead to a "coarsening of how people treat each other." In the future, calling Cortana a b**** or telling Siri to go f*** itself could lead to interesting consequences, such as reduced response time.

https://www.hardocp.com/news/2017/08...ual_assistants

[ímpar]

Boas!

This is Machine Learning

Machine learning beat down to its absolutely lowest level of explanation possible.



https://www.hardocp.com/news/2017/08...chine_learning

[ímpar]

Boas!

DeepMind launches new research team to investigate AI ethics
The Google-owned company will publish research on the effects of AI on society

https://www.theverge.com/2017/10/4/1...research-group

[joao veiga]

Estão a avançar bem depressa

[ímpar]

Boas!

Google's new browser experiment lets you learn about basic AI

https://www.theverge.com/tldr/2017/1...-ai-experiment

[ímpar]

Boas!

Pornhub is using machine learning to automatically tag its 5 million videos
The company is starting with facial recognition, but wants its AI to recognize specific categories too
...
Price told VentureBeat: “Ultimately, we want to provide our fans with whatever it is they have a penchant for, and our new model will be able to surface more accurate results for them, which will hopefully keep them coming back for. The easier it is to navigate and access what they want, the happier they will be.”
...
https://www.theverge.com/2017/10/11/...-video-tagging

[ímpar]

Boas!

Why Google is Terrified About Alexa and What it's Doing in Response

According to a report last month from Consumer Intelligence Research Partners, Amazon (WARNING auto play video) controls more than three-quarters of the market for home automation devices and Google accounts for most of the rest. This isn't a position Google likes to find itself in and they plan on doing something about it. One recent change they made was to add reminders to its home assistant that lets consumers get voice prompts to walk their dog or take out the trash. They want to personalize their devices as much as possible. Furthermore, Google sits on a mountain of data and they plan to use it to compete better with Amazon.

https://www.hardocp.com/news/2017/10...ng_in_response

[ímpar]

Boas!

Inspired by brain’s visual cortex, new AI utterly wrecks CAPTCHA security
Recognizing surfaces and edges key to picking out distorted characters.

https://arstechnica.com/science/2017...tcha-security/

[ímpar]

Boas!

Google’s AI thinks this turtle looks like a gun, which is a problem
New research shows how machine vision systems of all kinds can be tricked into misidentifying 3D objects

https://www.theverge.com/2017/11/2/1...fle-3d-printed

[joao veiga]

Ui . Sendo assim está muito atrás daquilo que imaginava.

[ímpar]

Boas!

In Amazon’s game engine, voice actors can now be replaced with robots
Just after actor strike ends, Lumberyard update adds a full text-to-speech pipeline.

https://arstechnica.com/gaming/2017/...d-with-robots/

[ímpar]

Boas!

Don't want a shipment of free gold? Send the scam email to this AI instead

The organization recently developed an AI chatbot which goes by the name of Re:scam. This AI has the wonderful job of replying to, and completely wasting the time of, email scammers. It has multiple personalities, is constantly learning new phrases, and anyone who receives a new scam email can forward it to Re:scam for the bot to get to work, wasting the time of internet thieves.

https://www.neowin.net/news/dont-wan...his-ai-instead

[ímpar]

Boas!

How Artificial Intelligence Is Making Nuclear Reactors Safer

Artificial intelligence continues to benefit us everyday. I know Elon thinks it's going to kill us off, but for now it's making nuclear reactors safer by helping to detect cracks in nuclear reactors.

https://www.hardocp.com/news/2017/11...reactors_safer