Profile cover photo
Profile photo
Gideon Rosenblatt
Grounding Machines in Humanity
Grounding Machines in Humanity

Gideon's posts

Post is pinned.Post has attachment
The alphabet changed the way we think. Artificial intelligence will do the same.

My latest, on alphabets, memorization, and what it says about the way we will know things in the future.

Like much of my writing, this one was inspired by a conversation here on Google+. In this case, it was +Laura Gibbs, recommending the book I talk about in this piece and interesting additional thoughts from +Thomas Morffew, +Sowmyan Tirumurti, and +John Verdon.

#knowledge #artificialintelligence #alphabet

Post has attachment
Alexa as Dominating the Voice Platform?

I must admit that I've not been giving Amazon's Alexa sufficient attention. According to this Backchannel piece, Amazon is set to really own this space. Personally, I think this race is far from over. It's just getting started and Google has many, many tricks up its sleeve on this front.

#alexa #voice #VoiceComputing

Post has attachment

Remember Google Posts? They're Now Open to Organizations

In case you forgot, Google is allowing celebs to create "Posts" that can be posted directly to the search results pages.

As of today, Google is now opening up this functionality to organizations and is initially targeting movies, museums, sports teams, and sports leagues.

"We’ve opened up posting on Google to museums, sports teams, sports leagues, and movies in the US, and all of the above plus musicians in Brazil. We’re continuing to experiment with other categories and regions, so stay tuned."

Want to see an example? Search for "vanessa da mata" and look at the box on the right side of the page, titled "Vanessa da Mata on Google". If you click through to more, you get a stream of posts like this:

How much traction will this approach get? No clue, but with something like this, why in the world would you ever bother with a Google+ Page if you're an organization trying to reach people? I wouldn't.

It's unclear what the approval process really entails, but here's the link if you're curious:

#Google #Posts

Post has attachment
This brings me back to the question of what the tech plutocrats mean by freedom. Martin Luther King led the March on Washington for “jobs and Freedom.” It’s obvious now that the new freedom brought to us by the libertarian elite will not come with jobs. The fact that Facebook generates revenues of $8 billion with less than 9000 employees speaks volumes. Is Peter Thiel’s idea of corporations, free to reap monopoly profits free from government regulation, what we want for our country? Thiel’s icon Ayn Rand defines freedom as “to ask nothing. To expect nothing. To depend on nothing.” How far is this from Jefferson’s great inspiration, the Greek philosopher Epicurus, who defines the good life in these terms?

* The company of good friends.

* The freedom and autonomy to enjoy meaningful work.

* The willingness to live an examined life with a core faith or philosophy.

I worry that our universities are being turned into trade schools in the pursuit of the almighty tech dollar. Are we forsaking the humanities and a basic liberal arts education all in promise to prepare students for the shark tank that awaits them in Silicon Valley or on Wall Street? As I said at the outset, I have no answers, but another phrase from Dr. King’s sermon calls out to me: “Our scientific power has outrun our spiritual power. We have guided missiles and unguided men.”

Let us not assume that this technological revolution we are living through has but one inevitable outcome. History is made by man, not by corporations or machines. It is time to wake up and begin to think about a digital renaissance. As my colleague Ethan Zuckerman said, “It’s obvious now what we did was a fiasco, so let me remind you that what we wanted to do was something brave and noble.” Your generation does not need to surrender to some sort of techno-determinist future. Let’s try and “rewire” (Ethan’s term) the Internet.

Post has attachment
Humanity, what are we doing?

Post has attachment
Trump's polarity in opposing climate action is creating an equal and opposite reaction. A great read by +Alex Steffen​.

Post has shared content
Imagine the most moving song in the world. Machine learning is now listening to your brain waves to make music.

Lots of room here for good, but also for emotional manipulation. 
Machine learning devices are learning how to compose songs that can elicit specific emotions in us humans 

Post has attachment
"人工知能" is Japanese for "Artificial Intelligence."

人 = human.
工 = made.
知 = knowing.
能 = ability.

Language is awesome, isn't it?

Update note: I should have clarified -- these are Chinese characters, and mean AI in Chinese too.

Second update: Nope, not quite. See the comments below to see the subtle differences between the Japanese and Chinese characters for AI. 

Post has shared content
Deep Mind Teaches AI to Remember It's Own Lessons

The Deep Mind folks have another important breakthrough. This time, it's a neutral network that didn't completely trash what it learned from earlier tasks in order to learn the next task. This kind of "memory" is critical to its ability to move from bespoke applications that are finely tuned for just one thing, to more general purpose applications.

“We are still a really long way from general-purpose artificial intelligence and there are many research challenges left to solve,” he added. “One key part of the puzzle is building systems that can learn to tackle new tasks and challenges while retaining the abilities that they have already learnt. This research is an early step in that direction, and could in time help us build problem-solving systems that can learn more flexibly and efficiently.”


Great catch, +David Amerland​.
Artificial General Intelligence in the Horizon

Humans learn because we don't forget the lessons we've been through while learning. Deep in our minds different connections are activated and different centers of the brain talk to each other and this renders us intelligent in the general sense.

The phenomenon is called "transfer" having learnt to punch a heavy bag made of leather and hanging from the ceiling, for instance, has given me skills that enable me to punch a grown man whose squared up against me and is hitting me back (we call this boxing :) ). In my brain the dynamics of motion and the physics of impact that I acquired while punching the heavy bag are superimposed upon the situation with my opponent sufficiently well for me (and him) to be able to train to square off against each other.

Computers powered by AI algorithms do not work like that. While they can learn to be faster and perhaps more precise than me when it comes to punching the heavy bag, the moment they square up (if they could) against a human opponent they need to go back to the drawing board, have a huge number of real matches and then begin to win.

Transfer, in short doesn't happen, because they do not remember what they learnt from that previous situation. But what if they could? What if they could learn sufficiently well as to be able to remember how they learnt so they can apply the principles in the new dynamic?

That is the premise of this fresh breakthrough by Google's DeepMind. Thankfully they are not quite up to speed yet and humans have a significant advantage, at present, but the gap is narrowing. :) 
Animated Photo

Post has shared content
Pretty cool.

HT +Daniel Estrada
Joto: Connected Whiteboard Robot Draws, Writes What You Like
More info:

neat :) 
Animated Photo
Wait while more posts are being loaded