Vertical AI startups. A new age of marketing. Amazon + Whole Foods. https://cognitionx.com/news-briefing/.
Facebook and Google have both just announced the ways they are using machine learning (as well as human brute force) to remove terrorist content from their platforms, thus thwarting terrorism with technology.
They are both using image matching, language understanding, and cross-platform collaboration to combat terrorism. This is, of course, in combination with the human efforts which they are using whether that be through specialists or through partnerships with other organisations.
I sincerely hope these efforts will have a lasting impact.
Tabitha UntilTheBotsTakeOver Goldstaub
P.S. 1 day till CogX. Be there or be intelligence artificial.
Bradford Cross describes the recipe for success for Vertical AI startups. In order to achieve success, these startups must have full stack products, subject matter expertise, propriety data, and AI delivering core value.
He also takes a deep look into the hottest sectors and markets and the importance that must be placed on getting the timing right. As he says, “The right idea with the right team at the wrong time == the wrong idea.”
Azeem Azhar (curator of The Exponential View), writes about the effects of robots taking our jobs by taking on a journey through history and the latest and greatest innovations in artificial intelligence.
He argues that it is not as black and white as people make it seem. Looking at the implementation of automated teller machines, for example, a less gloomy and sensationalistic trend appears: the rising number of ATMs didn’t oust bank clerks completely. People were redirected to complete other tasks in banks that haven’t been automated. He says, “In the world of the future, automated perfection is going to be common. Machines will bake perfect cakes, perfectly schedule appointments and keep an eye on your house. What is going to be scarce is human imperfection.
The research of Ned Block, Professor of Philosophy at NYU, is at the center of the vibrant academic debate about the true nature of consciousness. His work often straddles the boundary of philosophy of mind and cutting-edge neuroscience research, focusing on the philosophical conclusions about consciousness to be drawn from such research results.
In this talk, he discusses the so-called “hard problem” of consciousness and encourages the audience to participate in recreations of a series of fun studies that investigate the nature of consciousness without requiring the subjects to report anything. He uses these results to illustrate his theory that it’s possible for a subject to have conscious experiences that the subject isn’t paying attention to. Finally, he concludes by explaining why the advancements of AI, applied to understanding human cognition, are unlikely to solve the hard problem of consciousness.
In Musk’s paper on humans’ journey to Mars, the focus is on affordability, as that is the primary factor in making life on Mars a reality. As Musk notes, “You cannot create a self-sustaining civilization if the ticket price is $10 billion per person.” In order for it to be viable, Musk asserts that the cost should be about $200,000—equivalent to the median price of a house in the United States. In the paper, Musk outlines the steps he considers essential to ensuring this relative affordability.
Deep learning yields great results across many fields, from speech recognition, image classification, to translation. But for each problem, getting a deep model to work well involves research into the architecture and a long period of tuning. In this paper, Google researchers present a single model that yields good results on a number of problems spanning multiple domains. In particular, this single model is trained concurrently on ImageNet, multiple translation tasks, image captioning (COCO dataset), a speech recognition corpus, and an English parsing task.
Their model architecture incorporates building blocks from multiple domains. It contains convolutional layers, an attention mechanism, and sparsely-gated layers. Each of these computational blocks is crucial for a subset of the tasks we train on. Interestingly, even if a block is not crucial for a task, we observe that adding it never hurts performance and in most cases improves it on all tasks.
Amazon is in the news (again!) for their recent acquisition of Whole Foods and the big question everyone is asking is “why“? Besides for the 400+ physical stores that Amazon is getting, they are (perhaps more importantly) gaining 400 sources of prime data on consumer behaviour.
As Tom Krazit from Geekwire said, “Whole Foods instantly gives Amazon a reliable source of the purchasing habits of well-off Americans, and that data can be used to train artificial intelligence models that will allow retailers to better predict demand and someday automate much of the labor involved in grocery retailing.”
Rod Banner, Agent of Change at 3LA.com, believes that the advertising space is not what it used to be with the rise of social media. He speaks about the need to include the latest innovations in AI, such as chat bots, to stay ahead of the curves and argues that “the time when a freshly appointed agency Creative Director was able to wave a wand of wonder and change the fortunes of a brand feels over to [him].” Come hear him at CogX exploring the Impact of AI on Mental Health.