Tech investor Marc Andreessen took the stage Tuesday at TechCrunch Disrupt in San Francisco. SAN FRANCISCO – For tech entrepreneurs, these are both the best of times and the gnarliest of times. But only standout ideas with either unique intellectual property or quantifiable traction are likely to benefit in the form of investments or acquisitions from this buoyant climate. "Only strong companies are getting acquired," Greylock Ventures' Josh Elman told attendees at TechCrunch Disrupt Tuesday, citing the 1 billion price tags paid for self-driving car startup Cruise Automation (General Motors) and Dollar Shave Club (Unilever). Elman might have added another recent purchase to the list, Microsoft's 26 billion spend for professional networking site LinkedIn, founded by Elman's fellow Greylock colleague and panelist, Reid Hoffman.

Amazon's aim is to have Alexa indistinguishable from a human voice. Dave Limp, Amazon's SVP of device and services business is standing in front of an image of the bridge of the starship Enterprise, explaining the inspiration for Amazon's surprise hit Echo device. "A lot of people wondered what was the inspiration for this vision, it really was this, this cultural icon it started here with the tap on the lapel to talk to the computer. And later in the Star Trek series you could be anywhere on the starship Enterprise and you could talk to the computer and she would respond quickly with an answer," he says. With a 100M investment fund and the opening up of cloud service APIs and an SDK, Alexa and the Echo could become the brains of your home automation and IoT network.

The stated purpose of the research was to warn privacy and security enthusiasts of the threat from machine learning and AI to be used as a tool for identification and data collection. However, the threat may not be as bad as they claim. Machine learning cannot reverse the pixelation and recreate images, so anyone worried about blurred pictures of them can rest vaguely easy, at least for now. It can't actually recreate the pictures. It is also only successful when identifying things it's been specifically trained to look for.

Consider your repetitive daily tasks. Do you, or your employees, follow a process to perform these? If so, then this job, or at least the repetitive part of it, will almost definitely be replaced by a robotic usurper. No industry will be unchanged from the "robot revolution". Disruption is already occurring in many of the commoditised functions of accounting and legal advice, for example.

Recurrent neural networks are one of the staples of deep learning, allowing neural networks to work with sequences of data like text, audio and video. They can be used to boil a sequence down into a high-level understanding, to annotate sequences, and even to generate new sequences from scratch! The basic RNN design struggles with longer sequences, but a special variant – "long short-term memory" networks – can even work with these. Such models have been found to be very powerful, achieving remarkable results in many tasks including translation, voice recognition, and image captioning. As a result, recurrent neural networks have become very widespread in the last few years. As this has happened, we've seen a growing number of attempts to augment RNNs with new properties. Individually, these techniques are all potent extensions of RNNs, but the really striking thing is that they can be combined together, and seem to just be points in a broader space.

Diane Greene, executive vice president of Google Cloud Enterprise, appeared on stage at TechCrunch Disrupt San Francisco with TechCrunch editor Matt Burns today, and talked about her role running Google's massive cloud business. She came on board last year when Google bought her startup, bebop Technologies for 348 million, and she immediately brought with her enterprise credibility. Her resume includes helping start VMware before selling the company to EMC for 635 million, then helping bring it public under the EMC federation in 2007. At VMware, she worked on the virtual machine technology that would become the basic building blocks for cloud infrastructure services. Today, with that depth of experience, she is trying to help Google catch market leaders Amazon Web Services and Microsoft Azure.

Smart technology has made life so much easier in recent years. It's made homes safer, it's made meals tastier, and it's made watches a whole lot more complicated. Unfortunately, there is no device that can work out for you. There's no app that can motivate you to run, and there's no wearable that will give you abs. Working out is still something you need to do on your own.

Few trends in technology have caused the level of panic and uncertainty in the job market as artificial intelligence (AI). The impending "robot revolution" has brought questions about what jobs, if any, will be replaced by bots, and when it will happen. While some have posited that robots will replace nearly all jobs, and free up humans to work on more creative endeavors, others have been more reserved in their predictions. A new report for Forrester Research claims that, by 2021, "intelligent agents and related robots" will only have eliminated 6% of jobs. "By 2021, AI within intelligent agents will evolve significantly beyond today's relatively simple machine learning and natural language processing (NLP)," the report said.

Like most sectors, the solar industry is rapidly embracing ways to analyze and crunch data in order to lower the cost of solar energy and to open up new markets for their technology. The rise of data tools--algorithms, machine learning, sensors--are driving investments in, and acquisitions of, solar startups, while entrepreneurs are launching new companies that are using data to solve various solar industry problems. Meanwhile, big companies are spending money on tracking, monitoring and evaluating data from solar projects worldwide, helping to lower the cost of generating energy from the sun. It shouldn't come as a surprise that the solar sector is the latest to embrace the value of data. Other traditionally non-digital sectors, like the auto industry, oil and gas, and agriculture are turning to managing data as a necessity to keep their technology competitive and their companies in business.

In this special guest feature, Rob Farber from TechEnablement writes that modernized code can deliver significant speedups on machine learning applications. Benchmarks, customer experiences, and the technical literature have shown that code modernization can greatly increase application performance on both Intel Xeon and Intel Xeon Phi processors. Colfax Research recently published a study showing that image tagging performance using the open source NeuralTalk2 software can be improved 28x on Intel Xeon processors and by over 55x on the latest Intel Xeon Phi processors (specifically an Intel Xeon Phi processor 7210). For the study, Colfax Research focused on modernizing the C-language Torch middleware while only one line was changed in the high-level Lua scripts. NeuralTalk2 uses machine learning algorithms to analyze real-life photographs of complex scenes and produce a correct textual description of the objects in the scene and relationships between them (e.g., "a cat is sitting on a couch", "woman is holding a cell phone in her hand", "a horse-drawn carriage is moving through a field", etc.) Captioned examples are show in the figure below.