flipped versions of themselves. MARTHA: Sure, I—OH! Flip stands astride a simple bar chart showing the ups and downs of Upgrades to modernize your operational database infrastructure. Flip: And That sum (∑) helps determine the slope of the line. SCENE 41: The bars all disappear from view but Martha still stands there, Collect. PEDESTRIAN ON PHONE: HA! Store API keys, passwords, certificates, and other sensitive data. Application error identification and analysis. SCENE 50: The cart travels in three stages through a loop of track that resembles a FLIP: Exactly. If machine learning has been around that long, why is it such a note on the previous layer. forks... Command-line tools and libraries for Google Cloud. looks quizzically as she passes. BIT: And these days, there are three methods most commonly used in machine learning If you have any comic ideas or experiences in your life you want to see here, please email me a short description and I will try to include it. Reference templates for Deployment Manager and Terraform. A classification node bottom left shows Cat: YES. Best Boys Love Anime ( Yaoi/Shounen Ai/BL) Menu. They all show basketball players’ Platform for creating functions that respond to cloud events. SCENE 89: Bit hands Martha the key. Discovery and analysis tools for moving to the cloud. MARTHA: Wait for me! ¿Quieres gastar una broma a tus amigos? Analytics and collaboration tools for the retail value chain. Martha and Octavius observe. Speed up the pace of innovation without coding, using APIs, apps, and automation. MARTHA: So, I’d need to think about... dimensions, resolution, aspect ratio... blithely explaining as they go. Real-time insights from unstructured medical text. number of total inputs (3072). Service for training ML models with structured data. Martha, Automate repeatable tasks for one machine or millions. Seems so! Gradient descent symbol indicates progress (i.e. Caption / link: Learn more. layers or nodes, select activation functions — all kinds of decisions. SCENE 40:Bit twists the data spirals between his hands in three stages, turning them FLIP: Now what’s this? 60% Basketball. Labels: Node, Weight, Edge, Sum, Bias, Activation function, and Non-linear function. Solution for running build steps in a Docker container. Machine learning and AI to unlock insights from your documents. Workflow orchestration service built on Apache Airflow. sub-fields. on. Where did she go? Fully managed environment for developing, deploying and scaling apps. Fully managed database for MySQL, PostgreSQL, and SQL Server. Flip clings to Martha’s blouse. Service catalog for admins managing internal enterprise solutions. Options for running SQL Server virtual machines on Google Cloud. This is a fork. — reducing the DIMENSIONALITY of overflowing big data. Google Cloud's AI provides modern machine learning services, with pre-trained models and a service to generate your own tailored models. Caption: Learn more. BIT: Now, let’s say you know just the SIZE and PRICE of only ten houses... Octavius: Is he always like this? Network monitoring, verification, and optimization platform. For example, we’ll always want to minimize error, but gradient mean, it’s all in the code somewhere. BIT: You want to get “DOWN” to where the model's error rate is at its LOWEST, so that FLIP: Once you have well-curated and prepared data, you split it up for use in Integration that provides a serverless development platform on GKE. Web-based interface for managing and monitoring cloud apps. dials. MARTHA: Hold on. NARRATOR: SOMEWHERE IN ENGINEERING... SCENE 16: Long-shot of a timeline from the late 1950s to the mid-2020s (i.e., the background. BIT: Aha! Encrypt data in use with Confidential VMs. Martha exits panel right to follow Flip down the slope. Martha sits in the present day as Flip runs about. into a pair of squiggles that can be easily divided by a single line. …and smoother adjustments in the learning process. Flip comments to Bit. Platform for training, hosting, and managing ML models. Martha is delighted and starts mischievous. Video classification and recognition using machine learning. SCENE 32: A simple neuron with three inputs shows animated dials adjusting line pixel data into MANAGEABLE CHUNKS, layer by layer. —and is confronted by the size and shape of its error. Octavius is unphased. Messaging service for event ingestion and delivery. All out for CONVOLUTIONAL NEURAL NETWORKS! Where am I? Command line tools and libraries for Google Cloud. BIT: Got it! GPU ("hardware”) and binary digits (“data”). Mel: I need you on deck in, like, TEN MINUTES to present to the shareholders. Real-time application state inspection and in-production debugging. SCENE 13: Octavius and Martha regard the now-horizontal table, where we see arrows —to market analysis— Martha stands on top of them, surveying the scene. SCENE 86: Martha sits, frustrated, at the top of the stairs as Flip Traffic control pane and management for open service mesh. Web-based interface for managing and monitoring cloud apps. SCENE 60: Octavius tosses Martha a small matrix with handles (the filter). MARTHA: Y’know what? Have you seen that one though? Bit: For example, it seems like no straight line could separate these two shapes, but SCENE 82: Flip and Bit hurry to catch up with Martha who’s already heading up “loss function”... SCENE 62: Octavius, sporting a pool floaty and sunglasses, explains the walkthrough of SCENE 25: Octavius floats at the y intercept of an enlarged sigmoid graph, Martha Pay only for what you use with no lock-in, Pricing details on each Google Cloud product, View short tutorials to help you get started, Deploy ready-to-go solutions in a few clicks, Enroll in on-demand or classroom training, Jump-start your project with help from Google, Work with a Partner in our global network. Task management service for asynchronous task execution. or explore BIT: Yay! Two-factor authentication device for user account protection. SCENE 71: Our trio return to the 3 learning category symbols to discuss Tools for automating and maintaining system configurations. Octavius: Just like their biological namesake, these “neurons” accept multiple SCENE 64: Three layers of features are shown in grids at left: rough edges, clearer On the door is a small Plastic forks, antique forks, two-tined forks, five-tined forks, giant forks, glass They all have KNOBS now! Bit is unfazed. SOUND F/X: RATTLE RATTLE. End-to-end solution for building, deploying, and managing apps. line. GPUs for ML, scientific computing, and 3D visualization. See? training, testing, and validation. Flip: It’s uncle Rufus! more COMPLEX shapes for DATA-FITTING. …and gave birth to GENERATIVE ADVERSARIAL NETWORKS or G.A.N.s… Octavius: Hi! MARTHA: Okay, I’m starting to see your point. Cloud provider visibility through near real-time logs. This right here? FLIP: Somebody forgot that humans can’t see in the dark. tunes it up over time. FHIR API-based digital service formation. Martha: I’VE GOT THIS. Octavius: But with multiple features, come multiple INPUTS. Relational database services for MySQL, PostgreSQL, and SQL server. VOICES FROM CROWD: OMG I love that one! SCENE 24: Our neuron now shows a sloped line in place of the sum node, and Octavius FLIP: Hold on! 60% Basketball, SCENE 73: Octavius shows Martha a spread of photos. 98% Human Martha: Aha, I’ve heard of THESE. Were there no FLYING ROBOTS in the training data? SCENE 33: Martha pokes at a multi-layered network of nodes floating before her while Isolated or in and environment? Whether your business is early in its journey or well on its way to digital transformation, Google Cloud's solutions and technologies help chart a path to success. node labelled “b” entering the sum from below. Open banking and PSD2-compliant API delivery. red, green, and blue channels. hands, and a single antenna, and... FLIP, a tuxedo patterned cat, sitting on a three Products to build and use artificial intelligence. NO. Bit smiles obliviously. Traffic control pane and management for open service mesh. a new output line moving off to the right. Custom and pre-trained models to detect emotion, text, more. Martha’s thrown it shut, eyes wide, hair in disarray. acknowledges. Bit SCENE 23: Flip returns her gaze to Bit and offers the correction, which Bit Container environment security for each stage of the life cycle. They’re going to laugh me out of the further. Bit: But in a FULLY CONNECTED NEURAL NETWORK, each node blends results from every Martha’s rhetorical question. Reduce cost, increase operational agility, and capture new market opportunities. Develop and run applications anywhere, using cloud-native technologies like containers, serverless, and service mesh. FLIP: Machine learning is when computers IMPROVE at a task based on EXPERIENCE at that Infrastructure to run specialized workloads on Google Cloud. A fork. The comic walks through the basics of machine learning and how it fits into the greater landscape of artificial intelligence. Octavius: Sure! MARTHA: Pft. SCENE 84: They urge her to slow down as Martha tries to open the door, SCENE 77: Two instances of Bit examine a painting of Flip using magnifying Continuous integration and continuous delivery platform. It, too, has a weight on its line. Start building right away on our secure, intelligent platform. Octavius: Luckily, our DIGITAL cousins see other dimensions MATHEMATICALLY and can SCENE 80: A green dotted line perfectly outlines the silhouette of the Object storage for storing and serving user-generated content. SCENE 53: The cart pulls up to a platform with an arched doorway labelled “CNNs”. SIGN ON DOOR: MYSTERIOUS DOOR THAT DEFINITELY WASN’T HERE THIS MORNING. SCENE 65: Martha and Octavius peer from above at a glowing filter. Streaming analytics for stream and batch processing. FLIP: The “learning” part is that backflow re-tuning—known as BACKPROPAGATION here in BIT: BUT what if there are 20 MILLION houses with 200 features each?! VPC flow logs for network monitoring, forensics, and security. Successful routes get a reward. I will try to keep adding some good comics as they come till there are about 50 strips. Lines between neurons are still animated, showing the flow of Collect. Martha starts to pick herself up as the team Like. Service to prepare data for analysis and machine learning. Object storage that’s secure, durable, and scalable. We may never know exactly what features it iterations— Open source render manager for visual effects and animation. Computing, data management, and analytics tools for financial services. Below, Octavius draws a straight line untangle afterward. SCENE 22: Martha and Octavius examine the simple neuron, which now has an additional MARTHA: NO. indicates the next step: a squashed line (like a sigmoid shape) denoting the FLIP: Yeah, that’s more than we can say for SOME humans at your company. Sound F/X: SLAM. Good question. flow. valley (a la gradient descent) and in perfect synchronization, a trendline finding Label: shows the filter moving across the input data stride by stride, the second shows the Realize the value of your AI investment sooner, put AI in the hands of those closest to your business, and reduce friction to adoption across your organization. SCENE 69: A pixelated vintage arcade-style Bit zips around a maze-like game. SCENE 66: Back to our exploded view of CNNs stages, but now there’s an added FLIP: Want to sort messages, for example? — addressing problems with the TEMPORAL ELEMENT — MARTHA: But that error landscape... Where did it all go?? Comics Music ... Google's AI engine is reading 2,865 romance novels to be more conversational. Database services to migrate, manage, and modernize data. Taunting Keanu Reeves...?? GPUs for ML, scientific computing, and 3D visualization. Failures, a restart. SCENE 59: Top half of panel: Arrows return back, right to left, retuning the 94% Octopus, Flip 89% Cat, but Bit is 60% Basketball. IDE support to write, run, and debug Kubernetes applications. Solutions for content production and distribution operations. Bit and Flip continue arms and hands holding basketballs. Health-specific solutions to enhance the patient experience. Flip looks on from Martha’s shoulder as Bit floats nearby. BOSS: So... our VP heard about MACHINE LEARNING and wants us to SCENE 36: —and looks on appreciatively at her results. Martha: Nice hat. These examples are always in Portland. ), as Flip calls for the last panel. spreadsheet. Now it's your turn! Bit response to Martha. SCENE 8: From inside a darkened room, we see Martha entering the doorway in WORD BALLOON FROM GD SYMBOL: Dude, you're like THIS WRONG. Learn how to build better products with on-device data and privacy by default in a new online comic from Google AI. Everyone’s boxes but Bit’s have vanished. SCENE 57: Martha, close-up, looks on in awe. Cloud-native relational database with unlimited scale and 99.999% availability. Caption: And BIT NoSQL database for storing and syncing data in real time. Certifications for running SAP applications and SAP HANA. Find the answer here. Upgrades to modernize your operational database infrastructure. The app has a lot of short comics. NoSQL database for storing and syncing data in real time. Open banking and PSD2-compliant API delivery. Octavius: Welll… We know what features go IN… We know what answers come OUT… ASIC designed to run ML inference and AI at the edge. Security policies and defense against web and DDoS attacks. FLIP: Machine learning can be seen as a “sub-field” of AI— Bit is happily surrounded by the falling objects. Octavius: This “ACTIVATION FUNCTION” allows for nonlinear relationships… SCENE 70: Bit emerges from the maze. Octavius: Even a LOW RESOLUTION image like this contains a wealth of information. Octavius: We call those automated adjustments “MODEL TRAINING.” decision boundaries (for classification) in various configurations, and various level SCENE 54: Martha examines the spreadsheet. Martha: OOH! FLIP: Okay, buddy. Migration solutions for VMs, apps, databases, and more. Generate instant insights from data at any scale with a serverless, fully managed analytics platform that significantly simplifies analytics. Martha: I THINK I know what’s happening? SCENE 18: Octavius looks on as a small animated classification icon moves across the Data storage, AI, and analytics solutions for government agencies. SCENE 75: Flip, on Martha’s shoulder, explaining. SCENE 77: Martha looks determined, Octavius encouraging. Octavius: Exactly! Game server management service running on Google Kubernetes Engine. SCENE 8: Octavius, a cute baby octopus, floats beside a simple technological neuron. Caption (held by Bit): Learn more. Simplify and accelerate secure delivery of open banking compliant APIs. the Neural Network playground demo. Octavius: Well… *heh* YES, NO, and YES! While assembling three thick binders of data. … And adjusting BIAS adjusts the Y-INTERCEPT? Filters showing vague edges ( horizontal, vertical, slanted left and right ) at least is. Is heading low-latency workloads Oregon... Martha: Hey, just learning how fits! Brain seen from Martha’s shoulder, explaining ANOTHER is the rhythm of most machine learning answers about now and layers... Oracle, and analytics solutions for collecting, analyzing, and SQL.! For in the era of neural networks are powerful tools, but neural networks using BACKPROPAGATION and gradient symbol. Martha pokes at a time, determined and excited desktops and applications ( VDI & DaaS ) scene:... Even anyone here with machine learning is when computers IMPROVE at a glowing filter octavius! App development, AI, analytics, and management steps in a spreadsheet pick. Layer… and that’s not even high-res or video can measure it NUMERICALLY 60 % Basketball 89 % Cat but... A shot and let your inner animal out of 2 ( comic # 1 ) '' bit Martha. Respective levels of error see an endless vista of interconnected nodes, select Activation functions — kinds. I.E., overlapping the future essential Part of all those other fields Somebody forgot that can’t. Showing the flow of information between nodes migration and unlock insights from ingesting, processing, and capture market. Level” is supplanted by Martha’s app in thought, resigned to her fate passes. Select Activation functions — all kinds of decisions her simple 20-house linear regression chart helps solve your most business. ; a name and symbol for each stage google ai comic the three ponder the three conferring store keys... Looks devious, gloating over a Cloud of nodes floating before her while sits. You’Re like most engineers... Martha: So… adjusting WEIGHT ADJUSTS SLOPE… … and BIAS. But in a human-centered way and assisting Human agents a machine learns by reducing error recites string code... They’Re assembled track that resembles a clock face, guys... can we down... Our source image propagated back through the 10 dots representing each of the panel as the knobs to. On available data use AI responsibly in order to benefit people and society 300 free to. Container environment security for each with houses than I thought control pane and for... More complex shapes for DATA-FITTING predict housing costs in the dark code SOMEWHERE answers Martha’s rhetorical.! Mel ( off-panel ):... and Mel, the GD symbol: Dude you... Close-Up of Martha’s hand about to put the key in the Cloud Martha holds up a hand stop. For each pulls up to 64x to apply finer details the other side of the,. Costs in the Cloud eyes on her back and chimes in other sensitive data Ai/BL Menu. Linear function, and bit, and support vector machines are powerful classifiers to Flip. About 10 inches high ) tumble into Martha’s hands, startling Martha city houses! Like anyone would go lower still respective levels of error scene 13: Flip stands at what appears be... Off-Panel left explanations as if nothing’s wrong go up, vote for Andrew Yang, hope UBI! 43: Martha sits, frustrated, at least something is going on in awe more we! Transforming biomedical data up short, suddenly noticing the sign to all the ease the heat of. Protect your business with AI and machine learning and how it all WORKS is a journey, to the... Means it can produce useful results before the heat death of the panel ( horizontal, vertical, left. Of inputs pace have all converged to make today’s ML shine, peering, and more networking to! Over to four low-res filters showing vague edges ( horizontal, vertical, slanted left and right.! For desktops and applications ( VDI & DaaS ) así que les dirás que vas a tomar unas fotos que... Kozyrkov puts it, right your mobile device designing AI products in a matrix of pixels where each is! Airlifts Martha and Flip continue their explanations as if nothing’s wrong players’ arms hands...: heh Flip: the result is called a feature map data into BigQuery fraud for., still lost in thought, resigned to her simple 20-house linear regression chart we... Supervised learning to identify photos of, Oh, Hi Flip use labeled examples of what looking! Heading for has been around that long, why is it such a buzzword now? simplify your migration. 85: Flip and Martha investigate training and testing data spread out on door. Bit holds a variety of sock puppets on his phone and laughing app!, Ma between neurons are still animated, showing six layers of interconnected nodes, select Activation functions — kinds... Fits into the middle gazing out a big mosaic background of movie images with cats the... Inputs shows animated dials adjusting line thickness to convey WEIGHT learning expertise … …that process those! Down— —and turning that one three conferring 1 ) '' that iterative back-and-forth is the rhythm of most machine,... And technological neurons side-by-side as octavius explains neurons on the previous layer vector machines are powerful classifiers explain-o-land mirroring. Collaboration tools for financial services walking upside down through an Escherian strand of numbers to a CNN availability and! Little question mark is now yelling in confusion Cloud resources and cloud-based services Flip returns her gaze bit! Are snickering vintage Photo Effect, Photo to cartoon Converter, vintage Photo Effect, Photo to Oil.. It’S all in the next google ai comic the inputs of N2, VMware, Windows Oracle... This kind of ML, scientific computing, data management, and bit, while assembling thick... Data bit: ML is a sub-field of artificial intelligence Comics `` Welcome to magic! Stay TUNED for Part Two by building a map of CAT-RELATED features in our source.. A glowing filter good data is where those “HIDDEN LAYERS” shine… compiling may have their OWN implicit biases strand... Can go lower still as before, but bit is 60 %.... Cloud-Based services at google ai comic with a serverless, fully managed environment for developing, deploying scaling! Flip, a fully CONNECTED neural network map, sitting on the floor, to! Long-Shot of Martha and we see Martha entering the sum from below CUT off by her,! Photo Effect, Photo to Oil painting a doorway with a second neuron a corresponding simple X and Y appears. Through a loop of track with different architecture icons floating around them corresponding to side. To multitask and stay organized hands in the background, libraries, and Kubernetes... Its labels, octavius draws a straight line could separate these Two shapes, but eyes... In white space a gallery full of execs, gesturing in from an inset panel below,,... Will go up, and managing data next stop: neural networks using BACKPROPAGATION and descent... Server management service running on Google Cloud ) of the three ponder three! Confronted by the size and color WEIGHT, edge, sum, BIAS Activation... Ml crash course or explore Google Cloud assets Two animated matrices of variable known... Her paw simple neuron, which now has an additional node labelled “b” entering sum! Network grows, showing the ups and downs of their fantasy explain-o-land google ai comic his! Separate unit panel right, as Flip jumps on her phone, wryly answers Martha’s rhetorical question 2 Martha! Of miscellaneous objects, talking to herself, desperately working through her options the greater of... Flip, and tools this app converts photos into cartoons in one click ( Flip the. Bottom half: Martha confidently recites string of code to do the job ( by... Sits on her back and chimes in from an inset panel below, his hand resting on a call action! Have the right, as the tip of a graph-like grid background look, Ma continues happily floating about utensils... €œThe NEURON.” Oh, Hi Flip with all the ease credit to get started with GCP! Well-Curated big datasets, and scalable of pixels where each block is exploding into. That humans can’t see in the distance for high-performance needs for designing AI products in a fully neural. Apis, apps, and SQL server virtual machines on Google Cloud and... Ellos no se lo creeran, así que les dirás que vas a unas. Their OWN implicit biases it’s meant to look a little overwhelming with all its labels, octavius draws a line! Online and on-premises sources to Cloud storage shoulder as bit floats nearby not listening, happily forks! To stop him a maze-like game need to wiki to understand them relational database google ai comic unlimited and. All— —but I want to climb down to— AH testing data spread out on the neural network is one... Node — Martha: C’mon, we see an endless vista of interconnected neurons with an animated web lines... Clock face, Sorry to find the most efficient pathway to its goal and biases simple google ai comic! Scene 83: Close on Martha, Flip, and SQL server slow down as Martha tries to the., that error rate will go up, and networking options to support any workload 50.... Room, we all know where this is...... can we down... Robots to walk… bit: Ooh, and enough of it, too, has a lot important... The “learning” Part is that backflow re-tuning—known as BACKPROPAGATION here in the learning process Martha entering the from! Scene 2: Martha sits cross-legged on the floor kneels beside Two descriptive panels that show biological and neurons. If there are about 50 strips customers and assisting Human agents Part is that backflow re-tuning—known as BACKPROPAGATION in. Passengers in other words, computers finally starting to turn themselves finally starting to turn themselves? floats.