Continuations

Google formed and then within a week dissolved an AI Ethics board, in the corporate equivalent of many celebrity weddings: ill conceived, but quickly ended. The calls for ethics in AI have been strong and understandable. AI is powerful technology that can and already has gone terribly wrong, as in the advertising and recommendation algorithms used by Facebook and Youtube. Now some of this stuff is obvious and there has been no lack of people pointing out the problems. If you train an algorith...

We are finding ourselves in a strange place this Labor Day: the impact of AI is simultaneously being overestimated and underestimated. There is still massive hype surrounding self driving cars, which will almost certainly take much longer to arrive than the public appears to believe. At the same time, there is mostly quiet but massive AI progress on tasks such as reading radiology images. The key thing to understand – if you want to form an opinion on AI progress – is to distinguish between “...

Today the team at Clarifai is releasing an amazing second version of their API that makes two breakthroughs available to every developer: custom training and fast visual search. The beauty of Clarifai’s API when it first launched was that you could be up and running in minutes and get great results with their models. But what if you wanted to train your own concepts? Well, if you were a bigger customer willing to pay, Clarifai would train a custom model for you. With the newly released versio...

Natural language is the general purpose “API” for humans and it seems it may become the same for software. Bots are all the rage on several platforms including Telegram and now Slack. Facebook is experimenting with M. There are bots to schedule meetings. Bots to order lunch. And so much more. Our analyst Jonathan has wrote a great post about this development. Many of these bots are being built by new startups that are raising venture capital. This is fantastic news with regard to advancing hu...

Google just acquired DNNresearch, a spinoff company from the University of Toronto led by Geoffrey Hinton. The DNN in the company name stands for deep neural networks. As I have been learning from Professor Hinton’s online course, these are networks with multiple “hidden” layers of neurons. A hidden layer sits between an input and output layer. Hidden layers are responsible for much of the power of modern neural networks and the breakthroughs that have occurred are how to train these hidden l...

I recently wrote that the reason we can now get some real AI results is that we have a large body of data available. One area in which this applies is visual search. There are already some fantastic examples of what can be done with a large number of images in the work that has been done on pulling images from Flickr to assemble detailed views. There are also a couple of visual search engines already out there, such as Riya. The great thing for teaching a lot of these algorithms is that there...

This is a continuation of yesterday’s post. Those of us who have been around computers for a long time have lived through many letdowns when it comes to AI. The promise was frequently grand (cf various Herbert Simon quotes) but the actual results were disappointing to say the least. But a funny thing has happened over the last decade on the web. The grand promises have (mostly) disappeared and yet the results are beginning to show up! We might not be noticing them because they are not even la...

Today our portfolio company Clarifai has announced its $30 million Series B financing led by Menlo Ventures. It constitutes an important milestone in the history of the New York City startup ecosystem: generous early backing for an ambitious technology vision. I have been around the NY technology and startup world since the late 90s. There were a lot of justified knocks against the efforts here. It was difficult to find engineers for startups. Well, Google helped solve that by locating a majo...

I don’t (yet) worry about a Terminator or Matrix like scenario where machines take over the world and subjugate humans. But it is useful to remind oneself of the rate of progress that has been taking place. I started programming computers in my early teens which is now almost 35 years ago. During that the first 25 of those years progress towards any kind of artificial intelligence was painfully slow. For instance, as late as 2004 when DARPA ran the first Grand Challenge it looked like we were...

There has been a fair bit of coverage of the potential dangers that an Artificial General Intelligence (AGI) might pose for humanity. Here some questions about AGI with my currently best answers. 1. What is the difference between AI and AGI? Artificial Intelligence (AI) can accomplish specific tasks that we historically had a hard time to get computers to do well, such as recognize faces, categorize an image, drive a car, make a robot walk, etc. AI is here today. Artificial General Intelligen...