If you are looking for an answer to the question What is Artificial Intelligence? and you only have a minute, then here's the definition the Association for the Advancement of Artificial Intelligence offers on its home page: "the scientific understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines."
However, if you are fortunate enough to have more than a minute, then please get ready to embark upon an exciting journey exploring AI (but beware, it could last a lifetime) …
More than a year after announcing plans to automate the feature engineering phase of artificial intelligence projects, Seattle-based startup Kaskada Inc. is bringing its first product to market. Kaskada says it aims to democratize feature engineering, an often laborious process that requires data scientists to select, clean and validate the data to be fed into machine learning training models prior to moving them into production. A model intended to predict housing prices, for example, would be feature engineered with predictor data such as the square footage of properties, number of bedrooms and location. The larger and more complete the training data set, the better the results. The resources required to collect data and move machine learning models into production can be so significant that the capabilities are out of reach of all but the largest companies.
It is researched that 46 percent of US citizens use voice assistants. Observing the strong presence of voice assistants, banks, financial, service and insurance (BFSI) firms have actively adopted enterprise voice assistants for both internal (employees) and external (customers) purposes. It is said that JP Morgan & Co is enabling its clients by allowing access to research and analytics reports through voice chatbots. Also, twelve thousand field agents to be powered by voice assistant's capabilities, states Mark Madgett, the New York Life Insurances VP. Users can inquire about their account balance, latest transactions, fixed deposits, recurring deposits, loan balance, etc.
Many times AI has been put on a pedestal as the future of x y & z, however, many seem to agree that education is a sector in particular which will see stark changes in both admin, teaching styles, personalisation and more. I had the pleasure of speaking to three individuals working in the field, including, Vinod Bakthavachalam, Senior Data Scientist at Coursera, Kian Katanforoosh, Lecturer at Stanford University & Sergey Karayev, Co-Founder and CTO of Gradescope. We began by having Sergey of Gradescope walk us through his product, which has been recently acquired by turnitin. The concept, it seemed was formed from the simple and widespread issue of both lack of consistency, lack of insight through time constraint and delayed feedback on academic work. Sergey found that scanning the papers onto an online interface when paired with a rubric can allow for accurate marking in seconds across several papers.
In order to do machine learning engineering, a model must first be deployed, in most cases as a prediction API. In order to make this API work in production, model serving infrastructure must first be built. This includes load balancing, scaling, monitoring, updating, and much more. At first glance, all of this work seems familiar. Web developers and DevOps engineers have been automating microservice infrastructure for years now.
A comprehensive list of top startup companies who are building quite a reputation in the tech domain through voice tech offerings. Voice AI has been around since IBM introduced it in 1961 through IBM Shoebox. It was the first digital speech recognition tool which at its time could recognize 16 words and 9 digits. Today, using voice AI, developers can train neural network models, create human like voices, chatbots and more. The voice AI tech startups space is booming and now encompasses various avenues such as voice analytics, speech recognition, artificial voice synthesis, voice transcription, voice recognition, among others.
How do you teach an AI to walk? Artificial Intelligence, as we typically use the term right now, means a computational system that learns through pattern-spotting and self-correction, so you don't so much teach it as create a setting in which it can teach itself. If you want an AI to walk, you provide a set of constraints -- gravity exists, bodies are made of connected parts, the ground pushes back when you push on it -- and give it a challenge, like moving a certain distance. Then you step back and let it learn, and often marvel at the results. A recent paper entitled "The Surprising Creativity of Digital Evolution," published by a conglomerate of European and North American researchers, is packed with technically correct AI-devised solutions to the locomotion problem that are also, by any traditional measure, wrong.
It's not enough to have a pressure cooker, you need an Instant Pot that's also a slow cooker, and a rice cooker, and a yogurt maker. Your video game console is also now a media center and live streaming platform. And if your printer doesn't also make copies and send faxes, then what are you even doing with your life? This obsession with do-it-all gadgets has even hit the world of music gear. While there were certainly earlier examples, it really started to take off in the '90s with the emergence of the groovebox.
When I joined Amazon in 1998, the company had a single U.S.-based website selling only books and running a monolithic C application on five servers, a handful of Berkeley DBs for key/value data, and a relational database. That database was called "ACB" which stood for "Amazon.Com Books," a name that failed to reflect the range of our ambition. In 2006, acmqueue published a conversation between Jim Gray and Werner Vogels, Amazon's CTO, in which Vogel explained that Amazon should be viewed not just as an online bookstore but as a technology company. In the intervening 14 years, Amazon's distributed systems, and the patterns used to build and operate them, have grown in influence. In this follow-up conversation, Vogel and I pay particular attention to the lessons to be learned from the evolution of a single distributed system--Simple Storage Service (S3)--that was publicly launched close to the time of that 2006 conversation. TOM KILLALEA: In your keynote at the AWS re:Invent conference in December 2019, you said that in March 2006 when it launched, S3 was made up of eight services, and by 2019 it was up to 262 services. As I sat there I thought that's a breathtaking number, and it struck me that very little has been written about how a large-scale, always-on service evolves over a very extended period of time. That is a journey that would be of great interest to our software practitioner community. This is evolution at a scale that is unseen and certainly hasn't been broadly discussed. WERNER VOGELS: I absolutely agree that this is unparalleled scale. Even today, even though there are Internet services these days that have reached incredible scale--I mean look at Zoom, for example [this interview took place over Zoom]--I think S3 is still two or three generations ahead of that. Because we started earlier; it's just a matter of time, and at the same time having a strict feedback loop with your customers that continuously evolves the service. Believe me, when we were designing it, when we were building it, I don't think that anyone anticipated the complexity of it eventually. I think what we did realize is that we would not be running the same architecture six months later, or a year later. So, I think one of the tenets up front was don't lock yourself into your architecture, because two or three orders of magnitude of scale and you will have to rethink it.
Imagine for a moment that a road is used only for a single car and driver. Everything is smooth and wonderful. Then you wake up from that utopian dream and remember that our road networks have multiple cars of varying sizes, from different manufacturers, each with a driver with unique behaviors behind the wheel. We quickly realize that traffic conventions and rules are in place to avoid complete and utter chaos. We believe with increasing robotic use cases in the public domain as we all do see, a similar parallel reality needs to be realized and we propose that RoMi-H, an open-source robot and infrastructure framework that simplifies cross fleet robot collaboration, is the way to achieve this coming reality!
Monoclonal antibodies are an important weapon in the battle against COVID-19. However, these large proteins are difficult to produce in the needed quantities and at low cost. Attention has turned to nanobodies, which are aptly named, single-domain antibodies that are easier to produce and have the potential to be administered by inhalation. Koenig et al. describe four nanobodies that bind to the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein and prevent infection of cells (see the Perspective by Saelens and Schepens). Structures show that the nanobodies target two distinct epitopes on the SARS-CoV-2 spike protein. Multivalent nanobodies neutralize virus much more potently than single nanobodies, and multivalent nanobodies that bind two epitopes prevent the emergence of viral escape mutants. Science , this issue p. [eabe6230]; see also p.  ### INTRODUCTION The global scale and rapid spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pose unprecedented challenges to society, health care systems, and science. In addition to effective and safe vaccines, passive immunization by antibody-related molecules offers an opportunity to harness the vertebrate immune system to fight viral infections in high-risk patients. Variable domains of heavy-chain–only antibodies (VHHs), also known as nanobodies, are suitable lead molecules in such efforts, as they are small, extremely stable, easy to engineer, and economic to produce in simple expression systems. ### RATIONALE We engineered improved multivalent nanobodies neutralizing SARS-CoV-2 on the basis of two principles: (i) detailed structural information of their epitopes and binding modes to the viral spike protein and (ii) mechanistic insights into viral fusion with cellular membranes catalyzed by the spike. ### RESULTS Nanobodies specific for the receptor binding domain (RBD) of SARS-CoV-2 spike were identified by phage display using nanobody libraries from an alpaca and a llama immunized with the RBD and inactivated virus. Four of the resulting nanobodies—VHHs E, U, V, and W—potently neutralize SARS-CoV-2 and SARS-CoV-2–pseudotyped vesicular stomatitis virus. X-ray crystallography revealed that the nanobodies bind to two distinct epitopes on the RBD, interfaces “E” and “UVW,” which can be synergistically targeted by combinations of nanobodies to inhibit infection. Cryo–electron microscopy (cryo-EM) of trimeric spike in complex with VHH E and VHH V revealed that VHH E stabilizes a conformation of the spike with all three RBDs in the “up” conformation (3-up), a state that is typically associated with activation by receptor binding. In line with this observation, we found that VHH E triggers the fusion activity of spike in the absence of the cognate receptor ACE2. VHH V, by contrast, stabilizes spike in a 2-up conformation and does not induce fusion. On the basis of the structural information, we designed bi- and trivalent nanobodies with improved neutralizing properties. VHH EEE most potently inhibited infection, did not activate fusion, and likely inactivated virions by outcompeting interaction of the virus with its receptor. Yet evolution experiments revealed emergence of escape mutants in the spike with single–amino acid changes that were completely insensitive to inhibition by VHH EEE. VHH VE also neutralized more efficiently than VHH E or VHH V alone; stabilized the 3-up conformation of spike, as determined by cryo-EM; and more strongly induced the spike fusogenic activity. We conclude that the premature activation of the fusion machinery on virions was an unexpected mechanism of neutralization, as enhanced neutralization could not be attributed simply to better blocking of virus-receptor interactions. Activation of spike in the absence of target membranes likely induces irreversible conformational changes to assume the energetically favorable postfusion conformation without catalyzing fusion per se. Simultaneous targeting of two independent epitopes by VHH VE largely prevented the emergence of resistant escape mutants in evolution experiments. ### CONCLUSION Our results demonstrate the strength of the modular combination of nanobodies for neutralization. Premature activation of spike by nanobodies reveals an unusual mode of neutralization and yields insights into the mechanism of fusion. ![Figure] Bivalent nanobodies neutralize by inducing postfusion conformation of the SARS-CoV-2 spike. On virions, SARS-CoV-2 spike trimers are mostly in an inactive configuration with all RBDs in the down conformation (left). Binding of bivalent nanobody VE stabilizes the spike in an active conformation with all RBDs up (middle), triggering premature induction of the postfusion conformation, which irreversibly inactivates the spike protein (right). The pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to spread, with devastating consequences. For passive immunization efforts, nanobodies have size and cost advantages over conventional antibodies. In this study, we generated four neutralizing nanobodies that target the receptor binding domain of the SARS-CoV-2 spike protein. We used x-ray crystallography and cryo–electron microscopy to define two distinct binding epitopes. On the basis of these structures, we engineered multivalent nanobodies with more than 100 times the neutralizing activity of monovalent nanobodies. Biparatopic nanobody fusions suppressed the emergence of escape mutants. Several nanobody constructs neutralized through receptor binding competition, whereas other monovalent and biparatopic nanobodies triggered aberrant activation of the spike fusion machinery. These premature conformational changes in the spike protein forestalled productive fusion and rendered the virions noninfectious. : /lookup/doi/10.1126/science.abe6230 : /lookup/doi/10.1126/science.abg2294 : pending:yes