Google Coversational AI System and AI Test Kitchen

Test Kitchen is interesting. 

Google Debuts LaMDA 2 Conversational AI System and AI Test Kitchen   By Jaime Hampton in Datanami

At its annual developer conference last week, Google announced the second release of its conversational AI system, LaMDA, which stands for Language Models for Dialogue Applications. Google says LaMDa 2 is a more finely tuned version of their original release that uses open domain dialogue technology for conversation.

LaMDA is built on Google’s own Transformer neural network, an open source architecture that the company debuted in 2017. LaMDA was trained with dialogue (a dataset of 1.56 trillion words from public web data and documents, to be exact) to allow it to have what the company calls “natural, sensible and specific conversations.” It was then fine-tuned to generate natural language responses to given contexts and then classify its own responses according to whether they are safe and high quality. (More specific training information can be found in this developer blog post.)

Google CEO Sundar Pichai demonstrates LaMDA. With its “Imagine It” capability, users can input ideas and the AI will generate descriptive responses and even follow-up questions, as seen here.

During his keynote, Google CEO Sundar Pichai queued up pre-made demonstrations of LaMDA 2 that he said were not specific products but “quick sketches” to illustrate LaMDA’s capabilities. There is a creative function called “Imagine It,” where users can input their own ideas and the AI will generate imaginative and relevant descriptions of the idea. He shows LaMDA answering the prompt “Imagine I’m at the deepest part of the ocean” with “You’re in the Mariana Trench, the deepest part of the ocean! The waves are crashing against the walls of your submarine. You are surrounded by total darkness.” This description is followed by a list of follow-up questions generated by LaMDA, such as “What kind of creatures live here?” Ideas can be related to a straightforward topic (like the given example of the Mariana Trench), or a more abstract idea (such as how being on a planet made of ice cream might feel). Another function shown is called “Talk About It,” and this allows users to ask LaMDA 2 questions about a specific subject matter, after which it will stay on-topic, no matter which follow-up questions are asked.

Perhaps the most interesting capability, “List It,” was demonstrated live, this time by Senior Director of Product Management Josh Woodward, who said: “List It explores whether LaMDA can take a complex goal or topic and break it down into relevant subtasks. It can help me figure out what I’m trying to do and generate useful ideas I might not have thought of.”

Woodward tells LaMDA he wants to plant a vegetable garden, and it instantly provides him with a detailed list of subtasks, including “Make a list of what to grow” and “Research what grows best in your area.” Each suggested subtask can be broken down into a list of more suggestions specific to the task. LaMDA can also generate tips, such as trying container growing if yard space is a concern. .... ' 

The Army Studies Real Time Conversation

Considerable, interesting piece on the topic. Leads to the open question about how we converse with our robot assistants.   And leads on to other kinds of hybrid, cooperative work.   And AI to provide useful information about the meaning of statements and commands in context.    Expect lots more in this space in the coming years.

Army researchers create pioneering approach to real-time conversational AI  by The Army Research Laboratory  in Techexplore.

Spoken dialogue is the most natural way for people to interact with complex autonomous agents such as robots. Future Army operational environments will require technology that allows artificial intelligent agents to understand and carry out commands and interact with them as teammates.

Researchers from the U.S. Army Combat Capabilities Development Command, known as DEVCOM, Army Research Laboratory and the University of Southern California's Institute for Creative Technologies, a Department of Defense-sponsored University Affiliated Research Center, created an approach to flexibly interpret and respond to Soldier intent derived from spoken dialogue with autonomous systems.

This technology is currently the primary component for dialogue processing for the lab's Joint Understanding and Dialogue Interface, or JUDI, system, a prototype that enables bi-directional conversational interactions between Soldiers and autonomous systems.

"We employed a statistical classification technique for enabling conversational AI using state-of-the-art natural language understanding and dialogue management technologies," said Army researcher Dr. Felix Gervits. "The statistical language classifier enables autonomous systems to interpret the intent of a Soldier by recognizing the purpose of the communication and performing actions to realize the underlying intent."

For example, he said, if a robot receives a command to "turn 45 degrees and send a picture," it could interpret the instruction and carry out the task.

To achieve this, the researchers trained their classifier on a labeled data set of human-robot dialogue generated during a collaborative search-and-rescue task. The classifier learned a mapping of verbal commands to responses and actions, allowing it to apply this knowledge to new commands and respond appropriately....  " 

Seeking Conversational Writing

We  want our AIs and Bots to speak  and converse understandably.  Engaging with us and our colleagues and customers.   Efficiently providing answers and teaching too.   So what are the basics in making that happen?    Ultimately an important idea, which we struggled with when creating our speaking brand equities.   Here is a good start.

In the Alexa Blog.

4 Conversational Writing Tips That Will Transform Your Customer Interactions  By Lydia Roth.

Lydia is Digital Marketing Manager for Alexa. She has 4 years of experience in content marketing and demand generation, and is passionate about helping others grow their marketing know-how.

As more conversational AI assistants, such as Alexa and Siri, become mainstream, brands are looking for ways to make their own conversational experiences shine in the midst of the competition. Today’s rapidly shifting conversational marketing landscape has seen many technological advances, but some companies still fall into the trap of bland, robotic script writing that leaves many customers and prospects confused.

The good news is that rigid, robotic conversational experiences can be avoided by understanding key principles behind conversational writing. The key is to simply make your writing flow in a natural manner.

In this article, we’ll discuss four expert tips for conversational writing. We’ll focus on how these principles apply to creating conversational experiences, however these tips will provide value for anyone creating content. If that sounds like you, read on.   ... ' 

Experts on Conversational Marketing Trends

Good thoughts on how conversation supports marketing and decisions.  Stats about the conversational AI spac ealso interesting.

5 Experts Weigh in on Conversational Marketing Trends to Watch  in the Alexa Blog

With the conversational AI space expected to reach $15.7 billion by 2024—up from $4.2 billion in 2019—there’s never been a better time to invest in your company’s conversational marketing efforts.

Conversational experiences as a whole are going to continue to evolve and become more sophisticated in the coming years. Because of that, bots will also evolve to have more advanced capabilities and take on bigger and more important roles in the lives of businesses and consumers.

To help illustrate this shift, we’ve enlisted the help of five experts to share their thoughts and opinions about what conversational marketing trends your business should be aware of in the near future.... " 

Example of Question Answering Application: Jarvis

Question Answering Applications 

Developing Question a Question Answer Application with NVIDIA Jarvis   By James Sohn | February 25, 2021  Tags: AI/Deep Learning, BERT, cloud computing, featured,

There is a high chance that you have asked your smart speaker a question like, “How tall is Mount Everest?” If you did, it probably said, “Mount Everest is 29,032 feet above sea level.” Have you ever wondered how it found an answer for you?

Question answering (QA) is loosely defined as a system consisting of information retrieval (IR) and natural language processing (NLP), which is concerned with answering questions posed by humans in a natural language. If you are not familiar with information retrieval, it is a technique to obtain relevant information to a query, from a pool of resources, webpages, or documents in the database, for example. The easiest way to understand the concept is the search engine that you use daily. 

You then need an NLP system to find an answer within the IR system that is relevant to the query. Although I just listed what you need for building a QA system, it is not a trivial task to build IR and NLP from scratch. Here’s how NVIDIA Jarvis makes it easy to develop a QA system.

Jarvis overview

NVIDIA Jarvis is a fully accelerated application framework for building multimodal conversational AI services that use an end-to-end deep learning pipeline. The Jarvis framework includes optimized services for speech, vision, and natural language understanding (NLU) tasks. In addition to providing several pretrained models for the entire pipeline of your conversational AI service, Javis is also architected for deployment at scale. In this post, I look closely into the QA function of Jarvis and how you can create your own QA application with it.  ... " 

Dubber in the IBM Voice Cloud for Conversation Value

 Ultimately voice is data, and that data needs to be converted into intelligence to create useful conversations of value.   With logic, learning and AI.  To create application solutions like sales, assistance, reporting, control, tracking,  customer experience and many many more.  Here Dubber has been integrated for Telecom in the IBM Cloud.  Delivering value from conversations.

Dubber "Best of Breed" UCR & Voice Intelligence Cloud selected for IBM Cloud for Telecommunications Services    in PRNewswire 

Integrated Cloud offering to unlock efficiency and new services across Service Provider Networks globally. Addresses critical customer challenges in compliance, customer experience and security at scale

MELBOURNE, Australia, Nov. 5, 2020 /PRNewswire/ -- Dubber Corporation Limited (ASX: DUB) (Dubber), has been chosen as the recording and data capture platform for IBM Cloud for Telecommunications. Dubber's Voice Intelligence Cloud will enable Service Providers to deliver next-generation Unified Call Recording and Voice AI Services on a unified cloud platform. IBM announced Dubber on the IBM Cloud for Telecommunications today, selecting 'best-of-breed' solutions proven in leading Service Provider environments, as critical elements of the solution.

The only Unified Call Recording™ (UCR) solution integrated and interoperable with IBM Cloud for Telecommunications, Dubber enables Service Providers to offer solutions critical to their customers addressing compliance mandates, improving sales and service performance and unlocking the possibilities contained within voice data.

"IBM has assembled world-class 'best of breed' services to form a compelling Cloud offering for Service Providers with Dubber as a core component and the only Australian based company involved in the global offering," said James Slaney, co-founder, Dubber.

Today's announcements furthers Dubber's leadership globally in UCR and voice data as a service, increasing its ability to offer new services to its more than 140 Service Provider customers globally - and, for IBM to do the same with its global network of Service Provider customers.

"In simple terms, IBM and Dubber enable service providers to differentiate and create value from the conversations occurring on the network by making call recording and voice data as a service ubiquitous and open. All conversations can be captured in the Cloud, and converted to AI-enriched data," added Slaney. "Dubber will make it easy to capture all calls - from traditional landlines to 5G services, and any end-point provisioned by the Service Provider."  ... " 

Google Duplex Books Haircuts.

Fascinated by 'booking' as a simplistic form of contextual conversational.   And it is interesting that Google Duplex took so long to get to this variant.  But is that significant?    I can see 'booking' being a script for many kinds of useful interactions.  Certainly as the start of such an interaction.   For example for expert interaction we required the filling out of a problem description form, to determine if an expert was necessary.   Simple example

Google Duplex Can Book Haircuts, 2 Years After Stage Demo  By Eric Hal Schwartz in Voicebot,ai

The Google Duplex voice AI service can now book haircuts for clients, a service Google demonstrated when it announced Duplex in 2018. Duplex uses the Google Assistant AI to call barbers and salons, setting up appointments for their client. Until now, Duplex had been limited to making restaurant reservations or getting store information like opening times on behalf of users.  ... " 

Conversing Between Soldiers, Robots

Will be interesting to see how contextual such conversation can be.  Would seem risk and embedded goals would also be important.   Following to see what more I can learn.

Army Research Enables Conversations Between Soldiers, Robots
U.S. Army Research Laboratory
July 27, 2020

Researchers from the U.S. Army Combat Capabilities Development Command's Army Research Laboratory (ARL) and the University of Southern California's Institute for Creative Technologies have developed the Joint Understanding and Dialogue Interface (JUDI) capability, enabling conversations between soldiers and autonomous systems. ARL's Matthew Marge said JUDI enables interactions in tactical operations in which verbal task instructions can be employed for command and control of a mobile robot, and allows such a robot to request clarification or provide status updates as tasks are completed. Said Marge, "JUDI's ability to leverage natural language will reduce the learning curve for soldiers who will need to control or team with robots, some of which may contribute different capabilities to a mission, like scouting or delivery of supplies."  ... 

Politeness of/to Machines

Do we expect the same politeness from machines?  Are we polite to them?   They can detect some aspects of our demeanor by the words we use.     Will we begin to hide that?    Or is it useful to adjust the conversation?

Could Your Computer Please Be More Polite? Thank You
Carnegie Mellon University School of Computer Science
By Byron Spice

Carnegie Mellon University researchers have developed an automated method for restructuring nonpolite directives or requests into more polite communications. The team generated a dataset of 1.39 million sentences labeled for politeness—derived from publicly available emails exchanged by employees of long-gone energy firm Enron—for experimental application. Analysis determined the frequency and distribution of words in polite and nonpolite sentences, and the researchers developed a "tag and generate" pipeline to conduct politeness transfers. Impolite or nonpolite words or phrases were tagged, then a text generator replaced each tagged item without altering the meaning of the sentence. The system was able to produce subtler and more realistic restructures over time, and the team has released the labeled dataset for use by others in the hope of encouraging further politeness research.

Can AI Understand Cause and Effect?

Key point is made,  in human conversations we continually make cause-effect assumptions and consideration.

What AI still can’t do in TechnologyReview
Artificial intelligence won’t be very smart if computers don’t grasp cause and effect. That’s something even humans have trouble with.

by Brian Bergstein

In less than a decade, computers have become extremely good at diagnosing diseases, translating languages, and transcribing speech. They can outplay humans at complicated strategy games, create photorealistic images, and suggest useful replies to your emails.

Yet despite these impressive achievements, artificial intelligence has glaring weaknesses.

Machine-learning systems can be duped or confounded by situations they haven’t seen before. A self-driving car gets flummoxed by a scenario that a human driver could handle easily. An AI system laboriously trained to carry out one task (identifying cats, say) has to be taught all over again to do something else (identifying dogs). In the process, it’s liable to lose some of the expertise it had in the original task. Computer scientists call this problem “catastrophic forgetting.”

These shortcomings have something in common: they exist because AI systems don’t understand causation. They see that some events are associated with other events, but they don’t ascertain which things directly make other things happen. It’s as if you knew that the presence of clouds made rain likelier, but you didn’t know clouds caused rain. .... "

Facebook Blender Chatbot Does Open Source

Anything that can make a chatbot more conversational, and have the ability to use and update context
continually is good. 

Facebook releases its 'Blender' chatbot as an open-source project
It could help tomorrow's AI converse more naturally with people

Andrew Tarantola, @terrortola in Engadget

The virtual assistants that inhabit our smartphones are helpful, sure, but they’re not going to pass the Turing test any time soon. They’re designed for understanding specific commands and actions like checking on restaurant reservations or getting updates on the weather, rather than, say, carrying on an in-depth conversation with a human. But chatbots could soon become far more loquacious thanks to Facebook, which this morning released a startlingly lifelike chatbot that it’s been developing, dubbed Blender, as an open-source resource for AI research.

Facebook has been pouring money and resources into its Natural Language Processing technologies for a few years now and those efforts appear to have paid off. The company claims that Blender is the single largest open-source chatbot created to date. It’s been trained on a whopping 9.4 billion parameters -- nearly 4x as many as Google’s Meena and more than 10x as many as the previous largest OS chatbot available on the internet.

“One of the recent findings in the area of NLP, and AI in general, has been that as you scale, as these neural network models larger and larger, they tend to perform better,” Stephen Roller, a research engineer at Facebook’s AI lab (FAIR), told Engadget. “We had a number of issues when we were trying to train this thing. When you start to get that large, these things no longer are able to fit on a single GPU anymore.”  .... "

See more on their Blended Talk Ability.

Why Remote Does not Feel the Same

Interesting thoughts, mostly about the micro aspects, useful to think about during your next remote meeting.  Like the thoughts about how conversations really work, which we will need to promote intelligent results. 

The Google Blog:  WORKING AT GOOGLE
The science of why remote meetings don't feel the same

Zachary Yorke, UX Researcher
Published Apr 24, 2020

As COVID-19 has pushed more  teams to work remotely, many of us are turning to video calls. And if you’ve ever been on a video call and wondered why it doesn’t feel quite the same as an in-person conversation, we have something in common. As a researcher at Google, it’s my job to dig into the science behind remote communication. Here are a few things I’ve discovered along the way. 

#1: Milliseconds matter. 
As a species, we’re hardwired for the fast-paced exchange of in-person conversation. Humans have spent about 70,000 years learning to communicate face-to-face, but video conferencing is only about 100 years old. When the sound from someone’s mouth doesn’t reach your ears until a half second later, you notice. That’s because we’re ingrained to avoid talking at the same time while minimizing silence between turns. A delay of five-tenths of a second (500 ms)—whether from laggy audio or fumbling for the unmute button—is more than double what we’re used to in-person. These delays mess with the fundamental turn-taking mechanics of our conversations. 

On your next video conference, pump the brakes on your speaking speed to avoid unintended interruptions. If it’s a smaller group, try staying unmuted to provide little bits of verbal feedback (“mmhmm,” “okay”) to show you’re actively listening.   .... "

Seeking Better Chatbots with Contextual Memory

Another example of looking at memory to establish context understanding for a chatbot

Deploying more conversational chatbots
Startup Posh has created chatbots that use “conversational memory” to have more natural exchanges.

Zach Winn | MIT News Office

The comedian Bill Burr has said he refuses to call into automated customer service lines for fear that, years later on his death bed, all he’ll be able to think about are the moments he wasted dealing with chatbots.

Indeed, the frustrating experience of trying to complete even the most straightforward task through an automated customer service line is enough to make anyone question the purpose of life.

Now the startup Posh is trying to make conversations with chatbots more natural and less maddening. It’s accomplishing this with an artificial intelligence-powered system that uses “conversational memory” to help users complete tasks.

“We noticed bots in general would take what the user said at face value, without connecting the dots of what was said before in the conversation,” says Posh co-founder and CEO Karan Kashyap ’17, SM ’17. “If you think about your conversations with humans, especially in places like banks with tellers or in customer service, what you said in the past is very important, so we focused on making bots more humanlike by giving them the ability to remember historical information in a conversation.”

Posh’s chatbots are currently used by over a dozen credit unions across voice- and text-based channels. The well-defined customer base has allowed the company to train its system on only the most relevant data, improving performance.

The founders plan to gradually partner with companies in other sectors to gather industry-specific data and expand the use of their system without compromising performance. Down the line, Kashyap and Posh co-founder and CTO Matt McEachern ’17, SM ’18 plan to provide their chatbots as a platform for developers to build on.   .... "

Apple to Improve Siri with Voysis?

Most voice assistants today do a rocky job of interpreting complex conversation and context.  Beyond just simple interpretation, but on to understanding.  Beyond just 'Do this' or 'Do That'.   The basis of conversation is useful and credible response.   I know, I use several different assistant versions in context every day. Will Voysis help?  They claim domain specific voice AI.   Is that sufficiently similar to conversation context specific?

Apple's latest acquisition could help Siri understand what you're saying in Engadget

Voysis focused on AI that could respond to natural language requests.

The battle between AI voice assistants continues to rage on, and now Apple has acquired a tech firm, Voysis, that is all about helping computers understand natural language. As reported by Bloomberg, the firm's now-deleted website said it could produce search results from phrases like "I need a new LED TV, my budget is $1,000."  ... '

See also an article in TechCrunch.

More on Google Meena Chatbot: Making Contextual Sense

More on the recently announced Google chatbot.  As I have said before, we often see intelligence in our day to day world as a conversation.   Person to person, person to machine,  person to assistant, even person to document.  We have adjust to different expectations based on  'who' is communicating.    But there is also the matter of context, if its not well understood the apparent intelligence can be poor.   Consider 'making sense' to be a primary measure of achieving a conversational goal.    Making sense, common or otherwise, needs a firm contextual basis.  Evidence here:

Artificial intelligence: Does another huge language model prove anything?  By Ben Dickson in Techtalks

This article is part of our reviews of AI research papers, a series of posts that explore the latest findings in artificial intelligence.

This week, Google introduced Meena, a chatbot that can “chat about… anything.” Meena is the latest of many efforts by large tech companies trying to solve one of the toughest challenges of artificial intelligence: language.

“Current open-domain chatbots have a critical flaw — they often don’t make sense. They sometimes say things that are inconsistent with what has been said so far, or lack common sense and basic knowledge about the world,” Google’s researcher wrote in a blog post.

They’re right. Making sense of language and engaging in conversations is one of the most complicated functions of the human brain. Until now, most efforts to create AI that can understand language, engage in meaningful conversations, and generate coherent excerpts of text have yielded poor-to-modest results.  .... " 

IBM Upgrades Debate AI Tool to Better Derive Evidence

Intriguing approach to mining information to support a goal directed conversation.   A key aspect to making conversational systems more powerful.  Note also the crowdsourcing integrated here to grade evidence.  Noting that the report here does not mention 'Watson', it seems IBM is using their AI trademark much less these days.

IBM's Debating AI Just Got a Lot Closer to Being a Useful Tool
By MIT Technology Review via CACM

The IBM Debater system taking part in a debate at the University of Cambridge last year.
IBM upgraded the neural networks used by its Project Debater system, to improve the quality of evidence the argument-mining system uncovers.

IBM upgraded the neural networks used by its Project Debater system to improve the quality of evidence the argument-mining system uncovers.

One new add-on for the debating system is BERT (Bidirectional Encoder Representations from Transformers), a network designed by Google for natural language processing and answering queries.

IBM Research scientists trained the AI on 400 million documents from the LexisNexis database, providing a natural language dataset of roughly 10 billion sentences; the researchers combined the dataset with claims about several hundred different topics, then had crowdsourced workers label the sentences based on the quality of their evidence for or against specific claims.

A supervised learning algorithm digested this data, allowing BERT to manage queries on a wide range of subjects and to yield more relevant sentences compared to previous systems.

Project Debater was 95% accurate for the top 50 sentences across 100 distinct topics, according to IBM researcher Noam Slonim,  ... " 

Bixby Views and Touch

Saw Bixby early on, was not impressed.   Is it now emerging?  Touch it says will be included, which I did not see.

Bixby Views is What Every Voice Developer Wants and Reinforces Samsung’s Biggest Differentiator
  By Bret Kensella,  by Voicebot.ai

Putting Bixby on other Samsung devices particularly the refrigerators…I think that is the competitive edge Samsung has over the other competitors…More often than not, consumers have a Samsung appliance inside their home. If they buy a new one that is Bixby enabled then they have the interface right on their device versus adding a plugin.

Developers Realize Zero UI is a Myth

The idea of Zero UI and its potential to reduce developer time has turned out to be a myth. Voice is a UI and it is complex to design it well. With that said, there was a hope that with voice apps at least developers could avoid the complexities of graphical user interfaces (GUI) that invariably need to be ported to dozens or hundreds of form factors. The thinking was that a speaker and microphone are essentially the same on all devices, so we have the result of “write once; publish everywhere.”

But, those voice-only experiences didn’t last for long. Soon, Amazon, Google, and other OEMs introduced smart displays and started demonstrating voice apps on smart TVs. Google insisted all Actions for Assistant at least render text on the screen.


More Tools and More Modes to Support

Amazon then introduced Alexa Presentation Language (APL) and Google launched Interactive Canvas and suddenly not only was there more work to do to “publish everywhere,” but also some new tools to learn. Multimodal means that developing for voice apps can actually be more complex because you must accommodate both a voice user interface (VUI) and a GUI. This isn’t how it was supposed to be. But, here were are.

Bixby is a bit different from its voice assistant peers because it assumes multimodal features from the start. Google Assistant does as well, but you can get away with voice and text and skip the visuals if need be. Bixby expects an all-of-the-above approach of voice, visual, touch, and more. That makes it very flexible, but also adds some work. Then you have that issue of optimizing for all of those Samsung devices. So, it’s not a surprise that so many developers at SDC 19 were more excited about Bixby Views than the other announcements. It will help them reach more devices and save time.... "

Advances in Conversational Search

Here a long time interest in intelligent conversation, and one of the most used uses of conversation on line these days is search.   Intelligent conversation gets us much closer to general AI.    Has this now been considerably improved?  Are we getting much closer, or is this one more demo that cannot be broadly delivered?   Note too how Google mentions people adapting to machines with 'keyword-ese'  and avoiding natural conversation. 

Google now understands more conversational search queries
The tech giant says it's one of the biggest Search updates in the product's history.
By Mariella Moon, @mariella_moon  in Engadget ..... 

Understanding searches better than ever before
Pandu Nayak in the Google Blog
Google Fellow and Vice President, Search

If there’s one thing I’ve learned over the 15 years working on Google Search, it’s that people’s curiosity is endless. We see billions of searches every day, and 15 percent of those queries are ones we haven’t seen before--so we’ve built ways to return results for queries we can’t anticipate.

When people like you or I come to Search, we aren’t always quite sure about the best way to formulate a query. We might not know the right words to use, or how to spell something, because often times, we come to Search looking to learn--we don’t necessarily have the knowledge to begin with. 

At its core, Search is about understanding language. It’s our job to figure out what you’re searching for and surface helpful information from the web, no matter how you spell or combine the words in your query. While we’ve continued to improve our language understanding capabilities over the years, we sometimes still don’t quite get it right, particularly with complex or conversational queries. In fact, that’s one of the reasons why people often use “keyword-ese,” typing strings of words that they think we’ll understand, but aren’t actually how they’d naturally ask a question. 

With the latest advancements from our research team in the science of language understanding--made possible by machine learning--we’re making a significant improvement to how we understand queries, representing the biggest leap forward in the past five years, and one of the biggest leaps forward in the history of Search.   .... " 

Detecting Frustration to Enhance Conversation

A better means giving feedback?  We do in human conversation,  in a two way or multi-way conversation we stop for questions,  notice frowns,  gestures,  complaints.    A perfect conversation would have each component perfectly understood, absorbed, and then adjusted to.  But its only rarely happens that way.     Right now assistants ask you if their answer helped,  but rarely then adapt a new answer if not.   That's where some real human intelligence would live.  And taking that further to save the questions + adapted answers for later use.   Suggest below that deep learning can move to this.

Amazon is Testing a Way to Make the Thing You Hate Most About Alexa Go Away  in Inc.com

Ever found yourself screaming at your smart speaker? That just might work next time.
What if Alexa could tell if you were frustrated and course correct? That's exactly the feature Amazon will start testing, the company just announced at its September Devices Event. Alexa will soon have "frustration detection." It detects when Alexa gets your requests wrong, then tries to get it right. Amazon will start testing the feature with music requests in 2020, then will roll it out to other tasks gradually. 

Just say, "No, Alexa."

The feature will only be turned on for music requests to start. If Alexa plays the wrong song (e.g what is definitely not beach sounds), you can say, "No, Alexa." She'll apologize and ask you to clarify.

Here's how Amazon described the feature on their blog:

As customers continue to use Alexa more often, they want her to be more conversational and can get frustrated when Alexa gets something wrong. To help with this, we developed a deep learning model to detect when customers are frustrated, not with the world around them, but with Alexa. And when she recognizes you're frustrated with her, Alexa can now try to adjust, just like you or I would do. ... " 

(Update) AI Explainability Toolkit Talk and Technology

From last weeks talk on the just released open source explainabilty toolkit.   This can be seen as a fundamental part of most conversations.   When we interact with colleagues or with professionals, and get recommendations, we often have to ask the question 'Why?'.  This is an attempt at preloading AI originating answers to that question, based on a number of common templates.

http://cognitive-science.info/wp-content/uploads/2019/09/AIX360-CSIG-V1-2019-09-05.pdf  (Slides)

http://cognitive-science.info/community/weekly-update/  Update: Recording: https://www.youtube.com/watch?v=Yn4yduyoQh4

http://aix360.mybluemix.net/   (Technical link, demos)

What does it take to trust AI decisions ? 
AI is now used in many high-stakes decision making applications.

Addressing:
Is it fair?  Is it easy to understand?  Did anyone tamper with it?  Is it accountable?  

Very good talk, lots of great progress shown here,  but still lots more to do.   Everyone doing serious work with AI systems should examine this work and see how their system could link to this capability.  And extend it.   More to follow.

IBM Research AI Explainability 360 Toolkit

By Vijay Arya, Rachel Bellamy, Pin-Yu Chen,Payel Das, Amit Dhurandhar, MaryJo Fitzgerald,Michael Hind, Samuel Hoffman,Stephanie Houde, Vera Liao, Ronny Luss,Sameep Mehta, Saska Mojsilovic, Sami Mourad,Pablo Pedemonte, John Richards,Prasanna Sattigeri, Moninder Singh,Karthikeyan Shanmugam, Kush Varshney,Dennis Wei, Yunfeng Zhang, Ramya Raghavendra ....