Koomey's law This law posits that the energy efficiency of computation doubles roughly every one-and-a-half years (see Figure 1–7). In other words, the energy necessary for the same amount of computation halves in that time span. To visualize the exponential impact this has, consider the face that a fully charged MacBook Air, when applying the energy efficiency of computation of 1992, would completely drain its battery in a mere 1.5 seconds. According to Koomey's law, the energy requirements for computation in embedded devices is shrinking to the point that harvesting the required energy from ambient sources like solar power and thermal energy should suffice to power the computation necessary in many applications. Metcalfe's law This law has nothing to do with chips, but all to do with connectivity. Formulated by Robert Metcalfe as he invented Ethernet, the law essentially states that the value of a network increases exponentially with regard to the number of its nodes (see Figure 1–8).

Amazon Alexa AIのNatural Language Understandingグループは "Multilingual Amazon SLURP (SLU resource package) for Slot Filling, Intent Classification, and Virtual-Assistant Evaluation"(MASSIVE)をリリースした。自然言語理解(

Let's take a detailed look. This is the most common form of AI that you'd find in the market now. These Artificial Intelligence systems are designed to solve one single problem and would be able to execute a single task really well. By definition, they have narrow capabilities, like recommending a product for an e-commerce user or predicting the weather. This is the only kind of Artificial Intelligence that exists today. They're able to come close to human functioning in very specific contexts, and even surpass them in many instances, but only excelling in very controlled environments with a limited set of parameters. AGI is still a theoretical concept. It's defined as AI which has a human-level of cognitive function, across a wide variety of domains such as language processing, image processing, computational functioning and reasoning and so on.

There is plenty of talk about artificial intelligence in the enterprise, but a lot of it is not very practical. That's because enterprises aren't equipped with an army of data scientists to build and train new AI models. And it's not just the lack of qualified data scientists -- AI breakthroughs require massive amounts of relevant, annotated data. That doesn't mean however, there is no place for AI in your enterprise innovation strategy. Savvy CIOs are using in-market models and APIs by commercial and industry leaders to solve well-defined use cases, bringing immediate, measurable value to the organization.

Amazon Alexa might use the voice of friends and family who are no longer alive in a future update. Amazon mentioned the features at its re:MARS conference Wednesday as a way to "make memories last." After listening to someone's voice for less than a minute, Alexa would be able to simulate that voice when speaking. A video of the feature depicted a child who asked to have their grandmother read them a story, and Alexa affirmed before changing her voice, according to Sky News. It's not clear how far the feature is in development or when it could be rolled out to Alexa voice assistants.

Amazon's Alexa might soon replicate the voice of family members - even if they're dead. The capability, unveiled at Amazon's Re:Mars conference in Las Vegas, is in development and would allow the virtual assistant to mimic the voice of a specific person based on a less than a minute of provided recording. Rohit Prasad, senior vice president and head scientist for Alexa, said at the event Wednesday that the desire behind the feature was to build greater trust in the interactions users have with Alexa by putting more "human attributes of empathy and affect." "These attributes have become even more important during the ongoing pandemic when so many of us have lost ones that we love," Prasad said. "While AI can't eliminate that pain of loss, it can definitely make their memories last."

Do you miss the sound of a dead relative's voice? Well fear not: Amazon unveiled a new feature in the works for its virtual assistant Alexa that can read aloud in a deceased loved one's voice based on a short recording of the person. "While AI can't eliminate that pain of loss, it can definitely make their memories last," said Rohit Prasad, senior vice president and head scientist for Alexa, on Wednesday at Amazon's re:MARS conference in Las Vegas. In a video played at the event, an Amazon Echo Dot is asked: "Alexa, can Grandma finish reading me'The Wizard of Oz'?" "Instead of Alexa's voice reading the book, it's the kid's grandma's voice," Prasad said. "We had to learn to produce a high quality voice with less than a minute of recording."

Amazon plans to let people turn their dead loved ones' voices into digital assistants, with the company promising the ability to "make the memories last". The company is developing technology that will allow its Alexa digital assistant to mimic the voice of anyone it hears from less than a minute of provided audio, Rohit Prasad, its senior vice-president and head scientist, said on Wednesday. He added that during the coronavirus paramedic "so many of us have lost someone we love". While no timescale was given for the launch of the feature, the underlying technology has existed for several years. The company gave a demonstration where the reanimated voice of an older woman was used to read her grandson a bedtime story, after he asked Alexa: "Can grandma finish reading me the Wizard of Oz?" Prasad said: "The way we made it happen is by framing the problem as a voice conversion task and not a speech generation path."

Recommender systems personalize content by recommending items to users. Item recommendation algorithms are evaluated by metrics that compare the positions of truly relevant items among the recommended items. To speed up the computation of metrics, recent work often uses sampled metrics where only a smaller set of random items and the relevant items are ranked. This paper investigates such sampled metrics in more detail and shows that they are inconsistent with their exact counterpart, in the sense that they do not persist relative statements, for example, recommender A is better than B, not even in expectation. Moreover, the smaller the sample size, the less difference there is between metrics, and for very small sample size, all metrics collapse to the AUC metric. We show that it is possible to improve the quality of the sampled metrics by applying a correction, obtained by minimizing different criteria. We conclude with an empirical evaluation of the naive sampled metrics and their corrected variants. To summarize, our work suggests that sampling should be avoided for metric calculation, however if an experimental study needs to sample, the proposed corrections can improve the quality of the estimate. Recommender systems are a key technology in online platforms for personalizing the selection of items that are shown to a user. Examples include recommending which products to buy, which videos to watch or which songs to play. Recommendations are typically user-dependent and often context-dependent. A key operation of recommender systems is to retrieve a ranked list of the best items for a user in a particular context.

In Spain, the Madrid Metro uses AI to monitor its network and reduce energy consumption by 25%. In the U.S., a beverage company uses AI to drive sales by analyzing retailers and markets. In Europe, an energy company trains its engineers and managers in a digital twin factory powered by AI. In the Middle East, a telco's AI-powered virtual assistant speaks to 1.65 million customers every month in different Arab dialects and English. Undoubtedly, AI is in full adoption around the world, with all industries recognizing it as the next big thing in tech.