What GPT-3 on Azure will mean for Microsoft and OpenAI
GPT-3, Microsoft’s flagship language model, will be available through the new Azure OpenAI Service, announced at the Microsoft Ignite conference this week. The new service will be an opportunity for the software giant to strengthen its hold on the new business applications taking shape around advances in natural language processing and generation. It will also have implications for OpenAI, which is becoming increasingly dependent and entrenched in the business goals of Microsoft.
Google Research: Self-supervised learning is a game-changer for medical imaging
Artificial intelligence researchers at Google suggest a new technique that uses self-supervised learning to train deep learning models for medical imaging. In medical settings, data labeling is a complicated and costly endeavor. Google is one of several organizations that has been exploring its use in medical imaging in recent years. Early results show that the technique can reduce the need for annotated data and improve the performance of deep learning in medical applications.
Multimodal Neurons in Artificial Neural Networks
OpenAI announced CLIP, a general-purpose vision system that matches the performance of a ResNet-50, but outperforms existing vision systems. One such neuron, for example, is a “Spider-Man” neuron that responds to images of a spider, an image of the text “spider” and the comic book character “Spider” either in costume or illustrated.
How AI can enhance the human experience
Every new technology brings with it questions of ethics and unintended consequences. But I contend technologies like artificial intelligence can enhance, rather than reduce, the human experience. It’s important to remember that humans are completely at the helm with how these technologies are implemented and when and where they are restrained. The answer is adding humanity back into the digital world. It’s about striking balance.
AI meets BI: Key capabilities to look for in a modern BI platform
New age of self-service BI platforms has emerged and democratized data analytics for all. With the customer at its heart, modern augmented BI platforms no longer require scripting/coding skills or the knowledge to build the back-end data models, empowering even laymen to harness the power of raw data. As a user, here are the top AI capabilities that you need to look for in BI software.
Stop Blaming Humans for Bias in AI
As AI becomes more pervasive, we’re seeing more examples of how AI delivers value but also can spread harm through bias. A recent study showed that broadly targeted ads on Facebook for supermarket cashier positions were shown to an audience of 85 percent women. Blaming data scientists for bias is not the right solution. Finding a way to combat bias at a systemic level is the answer.
DeepMind becomes profitable and more enmeshed in Google’s business
DeepMind, the UK-based AI lab that seeks to develop artificial general intelligence, has finally become profitable, according to its latest financial report. The company finished the fiscal year with a £44 million profit, up from a £477 million loss in 2019. It has raked in £826 million in revenue in 2020, more than three times the £265 million it filed in 2019. It is not clear which one of DeepMind’s ventures caused the spike in revenue.
OpenAI’s API Now Available with No Waitlist
Starting today, developers in supported countries can sign up and start experimenting with our API right away. Improvements to our API over the past year include the Instruct Series models that adhere better to human instructions, specialized endpoints for more truthful question-answering, and a free content filter to help developers mitigate abuse. As our systems evolve, we expect to continue streamlining the process for developers, refining our Usage Guidelines, and allowing even more use cases over time.
Common Misconceptions About Differential Privacy
Differential privacy is a formal distinction between your secrets and secrets about you. Algorithms typically achieve the differential privacy standard by some type of noise addition to mask the presence of any record. DP is a standard that algorithms must meet to ensure that the output of the algorithm, such as a count, should not depend too much on any singular record. It does not provide guarantees about other ways your secrets become public.
8 ways AI & automation can complement each other (and your organization)
Many people no doubt confuse the two, and that isn’t helped by the way the media often conflates the two. Automation involves the application of technologies for carrying out processes with minimal human intervention. Robotics and software are forms of automation, but not necessarily include AI. Some types of automation are entirely devoid of AI. But when AI is added to a solution, a human contributor or gatekeeper can be subtracted from the equation.
CLIP: Connecting Text and Images
Current vision models are good at one task and one task only, and require significant effort to adapt to a new task. By design, the network can be instructed in natural language to perform a great variety of classification benchmarks. This is a key change: by not directly optimizing for the benchmark’s performance, it becomes much more representative. The idea of zero-data learning dates back over a decade but until recently was mostly studied in computer vision as a way of generalizing to unseen objects.
7 Reasons Why Enterprise Hesitates to Integrate AI in Mobile Apps
Artificial intelligence (AI) has always been a polarizing subject and in the tech industry, there are staunch, opposing sides. Enterprise companies appear to be leaning against incorporating AI solutions into their mobile apps for several valid reasons. But before we discuss those points, what exactly is this AI and how does it work? Artificial intelligence is the engineering of human thought patterns into a machine, especially computer systems.
What’s missing from self-serve BI and what we can do about it
The ambition of self-serve BI is an understandable one: these tools aspire to open up vast troves of “insight†by making data accessible to everyone at a company. But the reality is often messier. Interpreting data correctly requires a number of specific and nuanced skills, including the ability to reason in a certain way. To identify a better approach, we need to take a step back and determine what problem is actually trying to be solved.