Most of the recent advances in AI depend on deep learning, which is the use of backpropagation to train neural nets with multiple layers ("deep" neural nets).

Neural nets consist of layers of nodes, with edges from each node to the nodes in the next layer. The first and last layers are input and output. The output layer might only have two nodes, representing true or false. Each node holds a value representing how excited it is. Each edge has a value representing strength of connection, which determines how much of the excitement passes through.

The edges in an untrained neural net start with random values. The training data consists of a series of samples that are already labeled. If the output is wrong, the edges are adjusted according to how much they contributed to the error. It's called backpropagation because it starts with the output nodes and works toward the input nodes.

Deep neural nets can be effective, but only for single specific tasks. And they need huge sets of training data. They can also be tricked rather easily. Worse, someone who has access to the net can discover ways of adding noise to images that will make the net "see" things that obviously aren't there.

A post on the Keras blog was talking about eventually using AI to generate computer programs to match certain specifications. Gruber is saying something very similar.

Brain Games persuasion experiment: A fake sales person utilizes the halo effect, expert bias, and artificial scarcity to to promote a worthless product to consumers.

Google's DeepDream neural network applied to Joy of Painting with Bob Ross. Very creepy.

Google Home talking personal assistant spouts audio ads. Not much of a surprise.

Great overview and commentary. However, I would have liked some more insight into the ethical ramifications and potential destructiveness of an ASI-system as demonstrated in the movie.

According to a 2015 report by Incapsula, 48.5% of all web traffic are by bots.

...

The majority of bots are "bad bots" - scrapers that are harvesting emails and looking for content to steal, DDoS bots, hacking tools that are scanning websites for security vulnerabilities, spammers trying to sell the latest diet pill, ad bots that are clicking on your advertisements, etc.

...

Content on websites such as dev.to are reposted elsewhere, word-for-word, by scrapers programmed by Black Hat SEO specialists.

...

However, a new breed of scrapers exist - intelligent scrapers. They can search websites for sentences containing certain keywords, and then rewrite those sentences using "article spinning" techniques.

The team on Google Translate has developed a neural network that can translate language pairs for which it has not been directly trained. "For example, if the neural network has been taught to translate between English and Japanese, and English and Korean, it can also translate between Japanese and Korean without first going through English."

White House report: Artificial Intelligence, Automation, and the Economy

“Resistance is futile” Personal-ized education

2013 article about Douglas Hofstadter, who has continued to pursue an understanding of the human mind through experiments with AI.

We should have control of the algorithms and data that guide our experiences online, and increasingly offline. Under our guidance, they can be powerful personal assistants.

Big business has been very militant about protecting their "intellectual property". Yet they regard every detail of our personal lives as theirs to collect and sell at whim. What a bunch of little darlings they are.

The story of Walter Pitts and Warren McCulloch, whose theory of brain function was followed by John von Neumann's design for the stored program computer.

https://twitter.com/open_ai

They're hiring: https://openai.com/about/

The best computer vision company in Brazil

https://github.com/rhiever/tpot TPOT (Tree-based Pipeline Optimization Tool) Built on numpy, scipy, pandas, scikit-learn, and deap.

First comment

TLDR: Thought experiment that, by knowing about it, you are contributing to humanity enslavement to a all powerful AI

I would agree that "the problem of indexing and finding relevant information is tougher today than ever before" ... and especially "on one's own hard disk".

Vannevar Bush recognized the problem of artificial systems of indexing long before McIlroy pulled this page from his typewriter in 1964, and here we are 50 years later using the same kind of filesystem indexing systems and wondering why it's harder than ever to find information on our own hard drives.

With the advent of Google Docs we're finally moving away from the archaic indexing mentioned here. The filesystem metaphor was simple and dominated how everyone manages their data-- which extended into how we developed web content, as well.

The declaration that Hierarchical File Systems are Dead has led to better systems of tagging and search, but we're still far from where we need to be since there is still a heavy focus on the document as a whole instead of also the content within the document.

The linearity of printed books is even more treacherously entrenched in our minds than the classification systems used by libraries to store those books.

One day maybe we'll liberate every piece of content from every layer of its concentric cages: artificial systems of indexing, books, web pages, paragraphs, even sentences and words themselves. Only then will we be able to re-dress those thoughts automatically into those familiar and comforting forms that keep our thoughts caged.