Before starting to develop an AI strategy, make sure your team understands the limits of what is reasonable today, as well as incremental improvements that might be overlooked. Focus should be on your LOB leaders who understand the business. Make sure they are also able to recognize AI opportunities.
I recently saw this chart from PricewaterhouseCoopers as part of their ‘2018 AI Predictions Report’. The statement was: ‘we effectively utilize all the data we capture to drive business value’ (strongly agree – ~2200 C-level and IT respondents in large and mid-size companies internationally).
Keep in mind that the survey was done with future applications of AI as its focus. What immediately struck me is that these responses are almost exactly the inverse of what we should expect. Read More
Monthly Archives: May 2019
A Framework for Intelligence and Cortical Function Based on Grid Cells in the Neocortex
How the neocortex works is a mystery. In this paper we propose a novel framework for understanding its function. Grid cells are neurons in the entorhinal cortex that represent the location of an animal in its environment. Recent evidence suggests that grid cell-like neurons may also be present in the neocortex. We propose that grid cells exist throughout the neocortex, in every region and in every cortical column. They define a location-based framework for how the neocortex functions. Whereas grid cells in the entorhinal cortex represent the location of one thing, the body relative to its environment, we propose that cortical grid cells simultaneously represent the location of many things. Cortical columns in somatosensory cortex track the location of tactile features relative to the object being touched and cortical columns in visual cortex track the location of visual features relative to the object being viewed. We propose that mechanisms in the entorhinal cortex and hippocampus that evolved for learning the structure of environments are now used by the neocortex to learn the structure of objects. Having a representation of location in each cortical column suggests mechanisms for how the neocortex represents object compositionality and object behaviors. It leads to the hypothesis that every part of the neocortex learns complete models of objects and that there are many models of each object distributed throughout the neocortex. The similarity of circuitry observed in all cortical regions is strong evidence that even high-level cognitive tasks are learned and represented in a location-based framework. Read More
Superconducting Optoelectronic Neurons V: Networks and Scaling
Networks of superconducting optoelectronic neurons are investigated for their near-term technological potential and long-term physical limitations. Networks with short average path length, high clustering coefficient, and power-law degree distribution are designed using a growth model that assigns connections between new and existing nodes based on spatial distance as well as degree of existing nodes. The network construction algorithm is scalable to arbitrary levels of network hierarchy and achieves systems with fractal spatial properties and efficient wiring. By modeling the physical size of superconducting optoelectronic neurons, we calculate the area of these networks. A system with 8100 neurons and 330,430 total synapses will fit on a 1 cm × 1 cm die. Systems of millions of neurons with hundreds of millions of synapses will fit on a 300 mm wafer. For multi-wafer assemblies, communication at light speed enables a neuronal pool the size of a large data center (105 m2 ) comprising 100 trillion neurons with coherent oscillations at 1 MHz. Assuming a power law frequency distribution, as is necessary for self-organized criticality, we calculate the power consumption of the networks. We find the use of single photons for communication and superconducting circuits for computation leads to power density low enough to be cooled by liquid 4He for networks of any scale. Read More
Inside the Controversial Company Helping China Control the Future of the Internet
Dennis Honrud’s family has been farming wheat in eastern Montana for three generations. Unashamedly old school, Honrud sows only half his 6,000 acres, leaving the rest fallow to avoid soil depletion. “There’s not many of us left,” he laments. Like many workers in the global economy, the 68-year-old needs to stay connected, in his case to monitor crop prices and weather updates from his green John Deere tractor. So he asked a telecom provider to put a cell tower in his backyard.
The Honrud property in Glasgow, Mont., is so remote that it wasn’t well covered by any of the big four American carriers–Verizon, AT&T, T-Mobile and Sprint. So Honrud turned to the local provider, Nemont Wireless, to install the tower. Today, cell service is pretty good. When the occasional car accident happens on the stretch of highway next to the Honrud farm, highway patrol officers no longer need to drive a mile to get a signal. Now they can place a call from the scene. If that hasn’t saved a life yet, “at some point in time it will,” Honrud says.
But there’s a problem. Like around a quarter of the smaller “tier 3” carriers catering to rural areas like Glasgow, Nemont uses equipment provided by Huawei, the world’s biggest telecommunications-equipment company. The Chinese firm generated a mind-boggling $107 billion in revenue last year, selling equipment to customers in 170 countries and regions around the world. It also may be the most controversial company in the world. Read More
The AI Roles Some Companies Forget to Fill
AI is almost everywhere in the news today, and the drive to create and implement AI solutions is creating an enormous talent gap. An estimated 80% of companies are already investing in AI and most are facing challenges hiring the capabilities they need to implement a useful AI application or product. It’s clear that there is an intensively competitive market for artificial intelligence and machine learning specialists. Many companies first attempt to hire Ph.D.-level data scientists with expertise in AI algorithms and “feature engineering.” Some analysts have even equated “AI talent” with such researchers.
However, AI talent goes far beyond machine learning Ph.D’s. Equally important and less understood are the set of talent issues emerging around AI product development and engineering. Most firms have not filled these roles, and their AI projects are suffering as a result. Read More
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models
Several recent works have shown how highly realistic human head images can be obtained by training convolutional neural networks to generate them. In order to create a personalized talking head model, these works require training on a large dataset of images of a single person. However, in many practical scenarios, such personalized talking head models need to be learned from a few image views of a person, potentially even a single image. Here, we present a system with such few-shot capability. It performs lengthy meta-learning on a large dataset of videos, and after that is able to frame few- and one-shot learning of neural talking head models of previously unseen people as adversarial training problems with high capacity generators and discriminators. Crucially, the system is able to initialize the parameters of both the generator and the discriminator in a person-specific way, so that training can be based on just a few images and done quickly, despite the need to tune tens of millions of parameters. We show that such an approach is able to learn highly realistic and personalized talking head models of new people and even portrait paintings. Read More
Mona Lisa frown: Machine learning brings old paintings and photos to life
Machine learning researchers have produced a system that can recreate lifelike motion from just a single frame of a person’s face, opening up the possibility of animating not just photos but also paintings. It’s not perfect, but when it works, it is — like much AI work these days — eerie and fascinating.
The model is documented in a paper published by Samsung AI Center, which you can read here on Arxiv. It’s a new method of applying facial landmarks on a source face — any talking head will do — to the facial data of a target face, making the target face do what the source face does. Read More
How single neurons and brain networks support spatial navigation
Spatial navigation is an essential cognitive function, which is frequently impaired in patients suffering from neurological and psychiatric disorders. Research groups worldwide have studied the neuronal basis of spatial navigation, and the activity of both individual nerve cells and large cell assemblies in the brain appear to play a crucial role in the process. However, the relationship between the behaviour of individual cells and the behaviour of large cell networks has for the most part remained unexplored.
Various theories on this topic were put forward by an international team in the journal “Trends in Cognitive Sciences” from 24 May 2019. The review article was jointly authored by Dr. Lukas Kunz from the University Medical Center in Freiburg, Professor Liang Wang from the Chinese Academy of Sciences in Beijing, and Professor Nikolai Axmacher from Ruhr-Universität Bochum, together with colleagues from Columbia University in New York. Read More
Understanding Fake Agile
I was recently asked by a major corporation to give a talk on “fake agile.” They wanted me to explain what it is, how to identify it and how to deal with it. The request led me to give some thought to the many varieties of the beast, the reasons for its emergence and the prospects of taming, containing it, or turning it into the real thing.
The request is understandable. Some instances of supposedly agile management have as much relation to real Agile as someone wearing flamenco costumes and talking about flamenco, without having mastered flamenco dance steps or displaying a feel or flair for flamenco music.
With the growing recognition that “Agile is eating the world,” surveys by Deloitte and McKinsey show that more than 90% of senior executives give high priority to becoming agile, while less than 10% see their firm as currently highly agile. The gap between aspiration and reality has led to a vast number of managers, consultants, and coaches claiming to be agile and offering to help firms become agile. Quite a few firms also have CEOs who are asking, “Why aren’t we agile?”
As a result, the term “agile” is often thrown around without any agreement as to its meaning. Read More
When AI Becomes a Part of Our Daily Lives
As we live longer and technology continues its rapid arc of development, we can imagine a future where machines will augment our human abilities and help us make better life choices, from health to wealth. Instead of conducting a question and answer with a device on the countertop, we will be able to converse naturally with our virtual assistant that is fully embedded in our physical environment. Through our dialogue and digital breadcrumbs, it will understand our life goals and aspirations, our obligations and limitations. It will seamlessly and automatically help us budget and save for different life events, so we can spend more time enjoying life’s moments.
While we can imagine this future, the technology itself is not without challenges — at least for now. The ability for artificial intelligence to understand the complexities and nuances of human conversation is one hurdle. There are more than 7,111 known living languages in the world today, according to Ethnologue. Adding to the intricacies are the varied ways words are shared and used across different cultures, including grammar and the level of education and style of the speakers. Google Duplex, the technology supporting Google Assistant, which places phone calls using a natural-sounding human voice instead of a robotic one, is an early attempt to address such challenges in human communications. But these are just initial whispers in voice AI’s long journey. Read More
Rick's Cafe AI 