Get the most recent info and news about Kandji on HackerNoon, where 10k+ technologists publish stories for 4M+ monthly readers.

Kandji, the Device Harmony platform for Apple devices in the enterprise, tears down the wall between IT and InfoSec teams to keep Apple users secure and productive. In the past, these teams pursued their mandates independently — InfoSec handled security, while IT managed user devices. But in a world where user security and productivity are C-level priorities, they need shared intelligence, automation and workflows to win. 

Kandji’s Device Harmony platform deeply connects IT and InfoSec capabilities that were previously spread out across five different product categories: Device Management, Vulnerability Management, Endpoint Detection and Response, Endpoint Visibility, and Endpoint Compliance. Kandji bridges these capabilities with shared intelligence, automation, and workflows designed for the cross-functional manner in which these teams now work. 

Visit Kandji.io to request your free trial today.

Kandji

kandji

He replaced two fullbacks with attackers Macoumba Kandji and Claudio Lopez, and brought on defender Julien Baudet to anchor the three-man backline. Kandji helped set up the Rapids' lone goal on ...

Bringing on Kandji, Lopez pays dividends for Rapids

Lauren J. Mapp is a reporter at The San Diego Union-Tribune covering family caregiving, elder care, Indigenous communities and East County. She is a 2019 graduate from San Diego State University ...

Lauren J. Mapp

Please purchase an Enhanced Subscription to continue reading. To continue, please log in, or sign up for a new account. We offer one free story view per month. If you ...

Sporting Kansas City blanks Rapids 2-0

Dicker Data's New Zealand arm will distribute Kandji, which provides Apple device management solutions, in the local market. Kandji solutions can be used to manage devices running Apple's macOS ...

Dicker Data NZ signs Apple device management provider Kandji

That’s why the cybersecurity industry’s efforts have been more pointed at Windows, rather than macOS and Linux. Enter Kandji, a security platform for Apple endpoint device management. At the 2023 Black Hat USA conference in Las Vegas, ConversingLabs ...

With growing threats to Apple devices, Kandji ramps up

HARRISON, N.J. -- Red Bulls’ forward Mac Kandji sat at the lounge in Red Bull Arena on Saturday afternoon during the US-England viewing party, smiling, posing for photos and signing everything ...

New York's Kandji recuperating from injury quicker than expected

AI-powered image generation models are revolutionizing the creative landscape. The Midjourney platform has been a key player in this innovative field with its text-driven image creation. However, its Discord-based interface presented some limitations for professional use. Let's take a look instead at a new AI model called Kandinsky 2.2, a more builder-friendly text-to-image model available via a versatile API. Unlike Midjourney, which operates through Discord, Kandinsky enables developers to integrate AI image generation into various programming languages such as Python, Node.js, and cURLs. This means that with just a few lines of code, Kandinsky can automate the process of image generation, making it a more efficient tool for creative professionals. And with the new v2.2 release Kandinsky's image quality has never been higher. or follow me on for more content like this! Subscribe Twitter Kandinsky 2.2 brings a new level of accessibility and flexibility to AI image generation. It seamlessly integrates with multiple programming languages and tools, offering a level of flexibility that surpasses the Midjourney platform. Moreover, Kandinsky's advanced diffusion techniques result in impressively photorealistic images. Its API-first approach makes it easier for professionals to incorporate AI-powered visualization into their existing tech stack. In this guide, we'll explore Kandinsky's potential for scalability, automation, and integration, and discuss how it can contribute to the future of creativity. Join us as we delve into the tools and techniques needed to incorporate stunning AI art into your products using this advanced AI assistant. Key Benefits of Kandinsky 2.2 - Kandinsky is fully open source. Use the code directly or access it via Replicate's flexible API. Open source - Integrate Kandinsky into your workflows in Python, Node.js, cURLs, and more through the Replicate API. API access - Tweak images programmatically by modifying text prompts in code for rapid iteration. Automation - Generate thousands of images with simple API calls. Create storyboards and visualize concepts at scale. Scalability - Incorporate Kandinsky into your own tools and products thanks to its API-first design. Custom integration - Get granular control over image properties like lighting and angle through text prompts. ControlNet - Understands prompts in English, Chinese, Japanese, Korean, French, and more. Multilingual - Crisp, detailed 1024x1024 images ready for any use case. High resolution - State-of-the-art diffusion techniques produce stunning, realistic images on par with Midjourney. Photorealism How Does Kandinsky Work? Kandinsky 2.2 is a text-to-image diffusion model that generates images from text prompts. It consists of several key components: Text Encoder: The text prompt is passed through an XLM-Roberta-Large-Vit-L-14 encoder to extract semantic features and encode the text into a latent space. This produces a text embedding vector. Image Encoder: A pre-trained CLIP-ViT-G model encodes images into the same latent space as the text embeddings. This allows matching between text and image representations. Diffusion Prior: A transformer maps between the text embedding latent space and the image embedding latent space. This establishes a diffusion prior that links text and images probabilistically. UNet: A 1.22B parameter Latent Diffusion UNet serves as the backbone network. It takes an image embedding as input and outputs image samples from noisy to clean through iterative denoising. ControlNet: An additional neural network that conditions image generation on auxiliary inputs like depth maps. This enables controllable image synthesis. MoVQ Encoder/Decoder: A discrete VAE that compresses image embeddings as discrete latent codes for more efficient sampling. During training, text-image pairs are encoded to linked embeddings. The diffusion UNet is trained to invert these embeddings back to images through denoising. For inference, the text is encoded to an embedding, mapped through the diffusion prior to an image embedding, compressed by MoVQ, and inverted by the UNet to generate images iteratively. The additional ControlNet allows controlling attributes like depth. Key Improvements Over Prior Versions of Kandinsky An example showing the evolution of Kandinsky from v2.0 to v2.1 to v2.2. The realism! The primary enhancements in Kandinsky 2.2 include: : One of the key upgrades is the integration of the CLIP-ViT-G image encoder. This upgrade significantly bolsters the model's ability to generate aesthetically pleasing images. By utilizing a more powerful image encoder, Kandinsky 2.2 can better interpret text descriptions and translate them into visually captivating images. New Image Encoder - CLIP-ViT-G : Kandinsky 2.2 introduces the ControlNet mechanism, a feature that allows for precise control over the image generation process. This addition enhances the accuracy and appeal of the generated outputs. With ControlNet, the model gains the capability to manipulate images based on text guidance, opening up new avenues for creative exploration. ControlNet Support How Can I Use Kandinsky to Create Images? Ready to start creating with this powerful AI model? Here's a step-by-step guide to using the Replicate API to interact with Kandinsky 2.2. At a high level, you'll need to: - Get your Replicate API key and authenticate in your environment. Authenticate - Pass your textual description in the parameter. You can specify it in multiple languages. Send a prompt prompt - Tweak image dimensions, number of outputs, etc. as needed. Refer to the for more details, or read on. Customize parameters model spec - Kandinsky 2.2 outputs a URL to the generated image. Download this image for use in your project. Process the response For convenience, you may also want to try out this to get a feel for the model's capabilities before working on your code. live demo Step-by-Step Guide to Using Kandinsky 2.2 via the Replicate API In this example, we'll use Node to work with the model. So, you'll need to first install the Node.js client. npm install replicate Then, copy your API token and set it as an environment variable: export REPLICATE_API_TOKEN=r8_************************************* Next, run the model using the Node.js script: import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

const output = await replicate.run(
  "ai-forever/kandinsky-2.2:ea1addaab376f4dc227f5368bbd8eff901820fd1cc14ed8cad63b29249e9d463",
  {
    input: {
      prompt: "A moss covered astronaut with a black background"
    }
  }
); You can also set up a webhook for predictions to receive updates when the process is complete. const prediction = await replicate.predictions.create({
  version: "ea1addaab376f4dc227f5368bbd8eff901820fd1cc14ed8cad63b29249e9d463",
  input: {
    prompt: "A moss covered astronaut with a black background"
  },
  webhook: "https://example.com/your-webhook",
  webhook_events_filter: ["completed"]
}); As you work this code into your application, you'll want to experiment with the model's parameters. Let's take a look at Kandinsky's inputs and outputs. Kandinsky 2.2's Inputs and Outputs The text prompt is the core input that guides Kandinsky's image generation. By tweaking your prompt, you can shape the output. - The textual description, like "An astronaut playing chess on Mars." This is required. Prompt - Specifies elements to exclude, like "no space helmet." Optional. Negative Prompt - Image dimensions in pixels, from 384 to 2048. The default is 512 x 512. Width and Height - Number of denoising steps during diffusion, higher is slower but potentially higher quality. The default is 75. Num Inference Steps - Number of images to generate per prompt, default is 1. Num Outputs - Integer seed for randomization. Leave blank for random. Seed Combining creative prompts with these tuning parameters allows you to dial in your perfect image. Kandinsky Model Outputs Kandinsky outputs one or more image URLs based on your inputs. The URLs point to 1024x1024 JPG images hosted on the backend. You can download these images to use in your creative projects. The number of outputs depends on the "num_outputs" parameter. The output format looks like this: {
  "type": "array",
  "items": {
    "type": "string",
    "format": "uri"
  },
  "title": "Output"
} By generating variations, you can pick the best result or find inspiring directions. What Kinds of Apps or Products Can I Build With Kandinsky? The ability to turn text into images is a remarkable innovation, and Kandinsky 2.2 is at the forefront of this technology. Let's explore some practical ways this model could be used. In design, for instance, the rapid conversion of textual ideas into visual concepts could significantly streamline the creative process. Rather than relying on lengthy discussions and manual sketches, designers could use Kandinsky to instantly visualize their ideas, speeding up client approvals and revisions. In education, the transformation of complex textual descriptions into visual diagrams could make learning more engaging and accessible. Teachers could illustrate challenging concepts on the fly, enhancing students' comprehension and interest in subjects like biology or physics. The world of film and web design could also benefit from Kandinsky 2.2. By turning written scripts and concepts into visuals, directors and designers can preview their work in real time. This immediate visualization could simplify the planning stage and foster collaboration between team members. Moreover, Kandinsky's ability to produce high-quality images might open doors for new forms of artistic expression and professional applications. From digital art galleries to print media, the potential uses are broad and exciting. But let's not lose sight of the practical limitations. While the concept is promising, real-world integration will face challenges, and the quality of generated images may vary or require human oversight. Like any emerging technology, Kandinsky 2.2 will likely need refinement and adaptation to meet your needs. Taking It Further - Discover Similar Models With AIModels.fyi AIModels.fyi is a valuable resource for discovering AI models tailored to specific creative needs. You can explore various types of models, compare them, and even sort by price. It's a free platform that offers digest emails to keep you informed about new models. To find similar models to Kandinsky-2.2: Visit . AIModels.fyi Use the search bar to enter a description of your use case. For example, " " or " " realistic portraits . high-quality text to image generator View the model cards for each model and choose the best one for your use case. Check out the model details page for each model and compare to find your favorites. Conclusion In this guide, we've explored the innovative capabilities of Kandinsky-2.2, a multilingual text-to-image latent diffusion model. From understanding its technical implementation to utilizing it through step-by-step instructions, you're now equipped to leverage the power of AI in your creative endeavors. Additionally, AIModels.fyi opens doors to a world of possibilities by helping you discover and compare similar models. Embrace the potential of AI-driven content creation and subscribe for more tutorials, updates, and inspiration on AIModels.fyi. Happy exploring and creating! or follow me on for more content like this! Subscribe Twitter Further Reading: Exploring AI Models and Applications For those intrigued by the capabilities of AI models and their diverse applications, here are some relevant articles that delve into various aspects of AI-powered content generation and manipulation: : Discover how the AI Logo Generator Erlich leverages AI to create unique and visually appealing logos, expanding your understanding of AI's creative potential. AI Logo Generator: Erlich : Uncover a comprehensive overview of the best upscaling AI models, providing insights into enhancing image resolution and quality. Best Upscalers : Explore a detailed guide on how to effectively upscale images using the Midjourney AI model, enriching your knowledge of image enhancement techniques. How to Upscale in Midjourney: A Step-by-Step Guide : Dive into the realm of image denoising and restoration using ScuNet GAN, gaining insights into preserving image quality over time. Say Goodbye to Image Noise: How to Enhance Old Images with ScuNet GAN : Learn how the Gfpgan AI model breathes new life into old photos, providing you with a beginner's guide to revitalizing cherished memories. Breathe New Life into Old Photos with AI: A Beginner's Guide to Gfpgan : Gain insights into the nuances of AI-based face restoration by comparing the Gfpgan and Codeformer models. Comparing Gfpgan and Codeformer: A Deep Dive into AI Face Restoration : See the best models from the Nightmare AI team. NightmareAI: AI Models at Their Best : Understand the nuances between the ESRGAN and Real-ESRGAN AI models, shedding light on super-resolution techniques. ESRGAN vs. Real-ESRGAN: From Theoretical to Real-World Super Resolution with AI : Compare the Real-ESRGAN and SwinIR models, gaining insights into their effectiveness in image restoration and upscaling. Real-ESRGAN vs. SwinIR: AI Models for Restoration and Upscaling Also published here

"It's Like if Midjourney Had an API" - Taking a Look at Kandinsky 2.2

"মিডজার্নির একটি এপিআই থাকলে মনে হয়" - ক্যান্ডিনস্কি 2.2 এর দিকে নজর দেওয়া

„Es ist, als ob Midjourney eine API hätte“ – Ein Blick auf Kandinsky 2.2

"Es como si Midjourney tuviera una API": una mirada a Kandinsky 2.2

"C'est comme si Midjourney avait une API" - Jetez un œil à Kandinsky 2.2

"यह वैसा ही है जैसे मिडजॉर्नी के पास एक एपीआई होता" - कैंडिंस्की 2.2 पर एक नज़र डालें

「Midjourney に API があったようなもの」 - Kandinsky 2.2 を見てみる

"Midjourney에 API가 있는 것과 같습니다" - Kandinsky 2.2 살펴보기

"É como se Midjourney tivesse uma API" - Dando uma olhada em Kandinsky 2.2

"Sanki Midjourney'in bir API'si Varmış Gibi" - Kandinsky 2.2'ye Bir Bakış

"Giống như nếu Midjourney có API" - Hãy xem Kandinsky 2.2

“就像《中途旅程》有一个 API”——看一看康定斯基 2.2

I Googled, “most popular idioms” and this rich snippet included these 24 idioms: As idoms are defined as “ group of words established by usage as having from those of the individual words," I thought it would be fun to challenge a machine to deduce and visually produce what humans mean by our idioms. , I typed these idioms into our AI image generation system using the . a meaning not deducible Within the HackerNoon story draft Kandinsky 2 model Moon Break a leg Bite the bullet Call it a day Beat around the bush Go down in flames It takes two to tango Actions speak louder than words Through thick and thin Jump on the bandwagon On cloud nine The Whole Nine Yards Speak of the devil Come rain or shine A penny for your thoughts Ignorance is bliss Throw caution to the wind As right as rain Better late than never Raining cats and dogs A blessing in disguise Sitting on the fence A dime a dozen Head in the Clouds I’m very encouraged by how quickly the indie models are catching up to the mainstream models. Within the , you can now prompt Kandinsky 2, Midjourney Diffusion, and Stable Diffusion 1.5-2.1 (with negative prompts too). Comment below if there’s another AI image generation model you’d like us to add to HackerNoon. HackerNoon story draft editor

Google's 24 Most Popular Idioms Depicted by the Kandinsky 2 AI Art Model

This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. Authors: (1) Minghao Yan, University of Wisconsin-Madison; (2) Hongyi Wang, Carnegie Mellon University; (3) Shivaram Venkataraman, myan@cs.wisc.edu. Table of Links Abstract & Introduction Motivation Opportunities Architecture Overview Proble Formulation: Two-Phase Tuning Modeling Workload Interference Experiments Conclusion & References A. Hardware Details B. Experimental Results C. Arithmetic Intensity D. Predictor Analysis 2 MOTIVATION Many deep neural networks have been deployed on edge devices to perform tasks such as image classification, object detection, and dialogue systems. Scenarios including smart home assistants (He et al., 2020), inventory and supply chain monitoring (jet), and autopilot (Gog et al., 2022) often use battery-based devices that contain GPUs to perform the aforementioned tasks. In these scenarios, pre-trained models are installed on the devices where the inference workload is deployed. Prior works have focused on optimizing the energy consumption of GPUs (Wang et al., 2020b; 2021; Tang et al., 2019; Strubell et al., 2019; Mei et al., 2017) in cloud scenarios (Qiao et al., 2021; Wan et al., 2020; Hodak et al., 2019) and training settings (Wang et al., 2020a; Peng et al., 2019; Kang et al., 2022). On-device inference workloads exhibit different characteristics and warrant separate attention. In this section, we outline previous efforts in optimizing ondevice neural network inference and discuss our approach to holistically optimize energy consumption. 2.1 On-device Neural Network Deployment Prior work in optimizing on-device neural network inference focuses on quantization (Kim et al., 2021; Banner et al., 2018; Courbariaux et al., 2015; 2014; Gholami et al., 2021), designing hardware-friendly network architectures (Xu et al., 2019; Lee et al., 2019; Sanh et al., 2019; Touvron et al., 2021; Howard et al., 2019), and leveraging hardware components specific to mobile settings, such as DSPs (Lane & Georgiev, 2015). Our work explores an orthogonal dimension and aims to answer a different question: Given a neural network to deploy on a specific device, how can we tune the device to reduce energy consumption? In our work, we focus on edge devices that contain CPUs, memory, and GPUs. These devices are generally more powerful than DSPs often found on mobile devices. One such example is the Nvidia Jetson series, which is capable of handling a wide array of applications, ranging from AI to robotics and embedded IoT solutions (jet). The devices also come with dynamic voltage and frequency scaling (DVFS) capabilities that allow for the optimization of power consumption and thermal management during complex computational tasks. The Jetson series features a unified memory shared by both CPU and GPUs. We refer to the operating frequency of CPU, GPU, and shared memory as CPU frequency, GPU frequency, and memory frequency in this paper. To understand the system requirements in edge NN inference, we next describe a case study of how NNs are deployed in an inventory management company. From our conversations, Company A works with Customer B to deploy neural networks on edge devices to optimize inventory management. To comply with regulations and protect privacy, data from each inventory site are required to be stored locally. The vast difference in the layout of the inventories makes it impossible to pre-train the model on data from every warehouse. Therefore, these devices come with a pre-trained model based on data from a small sample of inventories, which may have significantly different layouts and external environments compared to the actual deployment venue. Consequently, daily fine-tuning is required to enhance performance in the deployed sites, as the environment continually evolves. Similar arguments apply to smart home devices, where a model is pre-trained on selected properties, but the deployed households may be much more diverse. To address privacy concerns, on-device Case Study on Inventory Management: fine-tuning of neural networks is preferred, as it keeps sensitive data locally. Therefore, edge devices often need to run both inference and periodic fine-tuning. Combining multiple workloads on edge devices can lead to SLO violations due to interference and increased energy use. 2.2 Holistic Energy Consumption Optimization Some recent works have explored reducing energy consumption by optimizing for batch size and GPU maximum frequency (You et al., 2022; Nabavinejad et al., 2021; Komoda et al., 2013; Gu et al., 2023) and developing power models for modern GPUs (Kandiah et al., 2021; Hong & Kim, 2010; Arafa et al., 2020; Lowe-Power et al., 2020). In this work, we argue that other hardware components also cause energy inefficiency and require separate optimization. We perform a grid search over GPU, memory, and CPU frequencies and various batch sizes to examine the Pareto frontier of inference latency and energy consumption. Figure 2 shows the tradeoff between the per-query energy consumption and inference latency (normalized to the optimal latency) on Jetson TX2 and Jetson Orin. Each point in the figure represents the optimal configuration that we find through grid search under a given inference latency budget and batch size. As Figure 2 shows, the Pareto frontier is not smooth globally and is difficult to capture by a simple model, which warrants more sophisticated optimization techniques to quickly converge to a hardware configuration that lies on the Pareto Frontier (Censor, 1977). Zeus (You et al., 2022) attempts to reduce the energy consumption of neural network training by changing the GPU power limit and tuning training batch size. PolyThrottle also includes these two factors. In Zeus (You et al., 2022), the focus is on training workloads in data center settings, where batch size tuning helps achieve an accuracy threshold in an energy-efficient way. We include batch size as part of PolyThrottle as it provides a trade-off between inference latency and throughput. Our empirical evaluation reveals new avenues available for optimization which complicates the search space, as we describe next.

PolyThrottle: Energy-efficient Neural Network Inference on Edge Devices: Motivation

This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. Authors: (1) Minghao Yan, University of Wisconsin-Madison; (2) Hongyi Wang, Carnegie Mellon University; (3) Shivaram Venkataraman, myan@cs.wisc.edu. Table of Links Abstract & Introduction Motivation Opportunities Architecture Overview Proble Formulation: Two-Phase Tuning Modeling Workload Interference Experiments Conclusion & References A. Hardware Details B. Experimental Results C. Arithmetic Intensity D. Predictor Analysis 8 CONCLUSION In this work, we examine the unique characteristics of energy consumption in neural network inference, especially for edge devices. We identified unique tradeoffs and dimensions between energy consumption and inference latency SLOs and empirically demonstrated hidden components in optimizing energy consumption. We then propose an optimization framework that automatically and holistically tunes various hardware components to find a configuration aligned with the Pareto Frontier. We empirically verify the effectiveness and efficiency of PolyThrottle. PolyThrottle also adapts to the need for fine-tuning and proposes a simple performance prediction model to adaptively schedule finetuning requests while keeping the online inference workload under the inference latency SLO whenever possible. We hope our study sheds more light on the hidden dimension of NN energy optimization. REFERENCES Jetson partner solutions ebook. URL https://resources.nvidia.com/ en-us-jetson-success-stories/ jetson-partner-solutions-ebook?lx=XRDs_y. Twitter streaming traces, 2018. URL https://archive.org/details/ archiveteam-twitter-stream-201804. Alipourfard, O., Liu, H. H., Chen, J., Venkataraman, S., Yu, M., and Zhang, M. Cherrypick: Adaptively unearthing the best cloud configurations for big data analytics. In NSDI, volume 2, pp. 4–2, 2017. Anderson, T., Belay, A., Chowdhury, M., Cidon, A., and Zhang, I. Treehouse: A case for carbon-aware datacenter software. 2022. Anthony, L. F. W., Kanding, B., and Selvan, R. Carbontracker: Tracking and predicting the carbon footprint of training deep learning models. 2020. Arafa, Y., ElWazir, A., ElKanishy, A., Aly, Y., Elsayed, A., Badawy, A., Chennupati, G., Eidenbenz, S., and Santhi, N. Verified instruction-level energy consumption measurement for nvidia gpus. 2020. Bai, Z., Zhang, Z., Zhu, Y., and Jin, X. Pipeswitch: Fast pipelined context switching for deep learning applications. In Proceedings of the 14th USENIX Conference on Operating Systems Design and Implementation, pp. 499–514, 2020. Banner, R., Hubara, I., Hoffer, E., and Soudry, D. Scalable methods for 8-bit training of neural networks. Advances in neural information processing systems, 31, 2018. Brochu, E., Cora, V. M., and De Freitas, N. A tutorial on bayesian optimization of expensive cost functions, with application to active user modeling and hierarchical reinforcement learning. arXiv preprint arXiv:1012.2599, 2010. Cai, H., Wang, T., Wu, Z., Wang, K., Lin, J., and Han, S. On-device image classification with proxyless neural architecture search and quantization-aware fine-tuning. In Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops, pp. 0–0, 2019. Cai, H., Gan, C., Zhu, L., and Han, S. Tinytl: Reduce memory, not parameters for efficient on-device learning. Advances in Neural Information Processing Systems, 33: 11285–11297, 2020. Cao, Q., Balasubramanian, A., and Balasubramanian, N. Towards accurate and reliable energy measurement of nlp models. In Proceedings of SustaiNLP: Workshop on Simple and Efficient Natural Language Processing, 2020. Censor, Y. Pareto optimality in multiobjective problems. Applied Mathematics and Optimization, 4(1):41–59, 1977. Courbariaux, M., Bengio, Y., and David, J.-P. Training deep neural networks with low precision multiplications. arXiv preprint arXiv:1412.7024, 2014. Courbariaux, M., Bengio, Y., and David, J.-P. Binaryconnect: Training deep neural networks with binary weights during propagations. Advances in neural information processing systems, 28, 2015. Crankshaw, D., Wang, X., Zhou, G., Franklin, M. J., Gonzalez, J. E., and Stoica, I. Clipper: A low latency online prediction serving system. In NSDI, volume 17, pp. 613– 627, 2017. Devlin, J., Chang, M.-W., Lee, K., and Toutanova, K. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018. Gardner, J. R., Kusner, M. J., Xu, Z. E., Weinberger, K. Q., and Cunningham, J. P. Bayesian optimization with inequality constraints. In ICML, volume 2014, pp. 937–945, 2014. Gholami, A., Kim, S., Dong, Z., Yao, Z., Mahoney, M. W., and Keutzer, K. A survey of quantization methods for efficient neural network inference. arXiv preprint arXiv:2103.13630, 2021. Gog, I., Kalra, S., Schafhalter, P., Gonzalez, J. E., and Stoica, I. D3: a dynamic deadline-driven approach for building autonomous vehicles. In Proceedings of the Seventeenth European Conference on Computer Systems, pp. 453–471, 2022. Gu, D., Xie, X., Huang, G., Jin, X., and Liu, X. Energyefficient gpu clusters scheduling for deep learning. arXiv preprint arXiv:2304.06381, 2023. Gujarati, A., Karimi, R., Alzayat, S., Hao, W., Kaufmann, A., Vigfusson, Y., and Mace, J. Serving {DNNs} like clockwork: Performance predictability from the bottom up. In 14th USENIX Symposium on Operating Systems Design and Implementation (OSDI 20), pp. 443–462, 2020. Gupta, U., Kim, Y. G., Lee, S., Tse, J., Lee, H.-H. S., Wei, G.-Y., Brooks, D., and Wu, C.-J. Chasing carbon: The elusive environmental footprint of computing. IEEE Micro, 42(4):37–47, 2022. He, C., Li, S., So, J., Zeng, X., Zhang, M., Wang, H., Wang, X., Vepakomma, P., Singh, A., Qiu, H., et al. Fedml: A research library and benchmark for federated machine learning. arXiv preprint arXiv:2007.13518, 2020. Hodak, M., Gorkovenko, M., and Dholakia, A. Towards power efficiency in deep learning on data center hardware. In IEEE International Conference on Big Data, 2019. Hong, S. and Kim, H. An integrated gpu power and performance model. In ISCA, 2010. Howard, A., Sandler, M., Chu, G., Chen, L.-C., Chen, B., Tan, M., Wang, W., Zhu, Y., Pang, R., Vasudevan, V., et al. Searching for mobilenetv3. pp. 1314–1324, 2019. Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., and Adam, H. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861, 2017. Ivanov, A., Dryden, N., Ben-Nun, T., Li, S., and Hoefler, T. Data movement is all you need: A case study on optimizing transformers. Proceedings of Machine Learning and Systems, 3:711–732, 2021. Kandasamy, K., Vysyaraju, K. R., Neiswanger, W., Paria, B., Collins, C. R., Schneider, J., Poczos, B., and Xing, E. P. Tuning hyperparameters without grad students: Scalable and robust bayesian optimisation with dragonfly. The Journal of Machine Learning Research, 21(1):3098–3124, 2020. Kandiah, V., Peverelle, S., Khairy, M., Pan, J., Manjunath, A., Rogers, T. G., Aamodt, T. M., and Hardavellas, N. Accelwattch: A power modeling framework for modern gpus. In MICRO, 2021. Kang, D.-K., Lee, K.-B., and Kim, Y.-C. Cost efficient gpu cluster management for training and inference of deep learning. Energies, 15(2):474, 2022. Kim, S., Gholami, A., Yao, Z., Mahoney, M. W., and Keutzer, K. I-bert: Integer-only bert quantization. pp. 5506–5518. PMLR, 2021. Klein, A., Bartels, S., Falkner, S., Hennig, P., and Hutter, F. Towards efficient bayesian optimization for big data. In NIPS 2015 Bayesian Optimization Workshop, 2015. Komoda, T., Hayashi, S., Nakada, T., Miwa, S., and Nakamura, H. Power capping of cpu-gpu heterogeneous systems through coordinating dvfs and task mapping. In 2013 IEEE 31st International Conference on computer design (ICCD). IEEE, 2013. Lacoste, A., Luccioni, A., Schmidt, V., and Dandres, T. Quantifying the carbon emissions of machine learning. arXiv preprint arXiv:1910.09700, 2019. Lane, N. D. and Georgiev, P. Can deep learning revolutionize mobile sensing? In Proceedings of the 16th international workshop on mobile computing systems and applications, pp. 117–122, 2015. Lee, J., Chirkov, N., Ignasheva, E., Pisarchyk, Y., Shieh, M., Riccardi, F., Sarokin, R., Kulik, A., and Grundmann, M. On-device neural net inference with mobile gpus. arXiv preprint arXiv:1907.01989, 2019. Lowe-Power, J., Ahmad, A. M., Akram, A., Alian, M., Amslinger, R., Andreozzi, M., Armejach, A., Asmussen, N., Beckmann, B., Bharadwaj, S., et al. The gem5 simulator: Version 20.0+. arXiv preprint arXiv:2007.03152, 2020. Mei, X., Wang, Q., and Chu, X. A survey and measurement study of gpu dvfs on energy conservation. Digital Communications and Networks, 3(2):89–100, 2017. Nabavinejad, S. M., Reda, S., and Ebrahimi, M. Batchsizer: Power-performance tradeoff for dnn inference. In Proceedings of the 26th Asia and South Pacific Design Automation Conference, 2021. NVIDIA. Stream management, 2023a. URL https://docs.nvidia.com/cuda/ cuda-runtime-api. NVIDIA. Nvidia multi-instance gpu, 2023b. URL https://docs.nvidia.com/datacenter/ tesla/mig-user-guide/index.html. Peng, Y., Zhu, Y., Chen, Y., Bao, Y., Yi, B., Lan, C., Wu, C., and Guo, C. A generic communication scheduler for distributed dnn training acceleration. In SOSP, 2019. Qiao, A., Choe, S. K., Subramanya, S. J., Neiswanger, W., Ho, Q., Zhang, H., Ganger, G. R., and Xing, E. P. Pollux: Coadaptive cluster scheduling for goodput-optimized deep learning. In OSDI, 2021. Rajpurkar, P., Zhang, J., Lopyrev, K., and Liang, P. Squad: 100,000+ questions for machine comprehension of text. In EMNLP, 2016. Romero, F., Li, Q., Yadwadkar, N. J., and Kozyrakis, C. {INFaaS}: Automated model-less inference serving. In 2021 USENIX Annual Technical Conference (USENIX ATC 21), pp. 397–411, 2021. Sanh, V., Debut, L., Chaumond, J., and Wolf, T. Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter. arXiv preprint arXiv:1910.01108, 2019. Schwartz, R., Dodge, J., Smith, N. A., and Etzioni, O. Green ai. Commun. ACM, 63(12):54–63, 2020. Shen, H., Chen, L., Jin, Y., Zhao, L., Kong, B., Philipose, M., Krishnamurthy, A., and Sundaram, R. Nexus: A gpu cluster engine for accelerating dnn-based video analysis. In Proceedings of the 27th ACM Symposium on Operating Systems Principles, pp. 322–337, 2019. Snoek, J., Larochelle, H., and Adams, R. P. Practical bayesian optimization of machine learning algorithms. Advances in neural information processing systems, 25, 2012. Strubell, E., Ganesh, A., and McCallum, A. Energy and policy considerations for deep learning in nlp. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 2019. Tambe, T., Hooper, C., Pentecost, L., Jia, T., Yang, E.- Y., Donato, M., Sanh, V., Whatmough, P., Rush, A. M., Brooks, D., et al. Edgebert: Sentence-level energy optimizations for latency-aware multi-task nlp inference. In MICRO, 2021. Tan, M. and Le, Q. Efficientnet: Rethinking model scaling for convolutional neural networks. pp.6105–6114. PMLR, 2019. Tang, Z., Wang, Y., Wang, Q., and Chu, X. The impact of gpu dvfs on the energy and performance of deep learning: An empirical study. In Proceedings of the Tenth ACM International Conference on Future Energy Systems, 2019. Touvron, H., Cord, M., Douze, M., Massa, F., Sablayrolles, A., and Jegou, H. Training data-efficient image transform- ´ ers & distillation through attention. pp. 10347–10357. PMLR, 2021. Venkataraman, S., Yang, Z., Franklin, M., Recht, B., and Stoica, I. Ernest: Efficient performance prediction for {Large-Scale} advanced analytics. In 13th USENIX Symposium on Networked Systems Design and Implementation (NSDI 16), pp. 363–378, 2016. Wan, C., Santriaji, M., Rogers, E., Hoffmann, H., Maire, M., and Lu, S. Alert: Accurate learning for energy and timeliness. In ATC, 2020. Wang, F., Zhang, W., Lai, S., Hao, M., and Wang, Z. Dynamic gpu energy optimization for machine learning training workloads. IEEE Transactions on Parallel and Distributed Systems, 2021. Wang, G., Venkataraman, S., Phanishayee, A., Devanur, N., Thelin, J., and Stoica, I. Blink: Fast and generic collectives for distributed ml. In Proceedings of Machine Learning and Systems, 2020a. Wang, Y., Wang, Q., Shi, S., He, X., Tang, Z., Zhao, K., and Chu, X. Benchmarking the performance and energy efficiency of ai accelerators for ai training. In 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGRID), 2020b. Wolf, T., Debut, L., Sanh, V., Chaumond, J., Delangue, C., Moi, A., Cistac, P., Rault, T., Louf, R., Funtowicz, M., Davison, J., Shleifer, S., von Platen, P., Ma, C., Jernite, Y., Plu, J., Xu, C., Scao, T. L., Gugger, S., Drame, M., Lhoest, Q., and Rush, A. Transformers: State-of-the-art natural language processing. In EMNLP, 2020. Wu, C.-J., Raghavendra, R., Gupta, U., Acun, B., Ardalani, N., Maeng, K., Chang, G., Aga, F., Huang, J., Bai, C., Gschwind, M., Gupta, A., Ott, M., Melnikov, A., Candido, S., Brooks, D., Chauhan, G., Lee, B.,Lee, H.-H., Akyildiz, B., Balandat, M., Spisak, J., Jain, R., Rabbat, M., and Hazelwood, K. Sustainable ai: Environmental implications, challenges and opportunities. In Proceedings of Machine Learning and Systems, 2022. Wu, X., Rao, J., Chen, W., Huang, H., Ding, C., and Huang, H. Switchflow: preemptive multitasking for deep learning. In Proceedings of the 22nd Inter Xu, M., Liu, J., Liu, Y., Lin, F. X., Liu, Y., and Liu, X. A first look at deep learning apps on smartphones. In The World Wide Web Conference, WWW ’19, pp. 2125–2136, 2019. You, J., Chung, J.-W., and Chowdhury, M. Zeus: Understanding and optimizing gpu energy consumption of dnn training. arXiv preprint arXiv:2208.06102, 2022. Yu, P. and Chowdhury, M. Fine-grained gpu sharing primitives for deep learning applications. Proceedings of Machine Learning and Systems, 2:98–111, 2020. Zhao, Y., Liu, X., Liu, S., Li, X., Zhu, Y., Huang, G., Liu, X., and Jin, X. Muxflow: Efficient and safe gpu sharing in large-scale production deep learning clusters. arXiv preprint arXiv:2303.13803, 2023.

PolyThrottle: Energy-efficient Neural Network Inference on Edge Devices: Conclusion & References

Software Engineering for OpenHarmony—A Research Roadmap: Conclusion & References

1. We’ve Already Been Trained to Post What Where Adding social to media, making social media, turned the documenter into the subject matter, as we were then all expected to become sources. We turned out to be social media armchair experts, posting more about other people, their likeness, and world events. All these updates show off more about our own existing perspectives. Large Language Models (LLMs) have so much documentation by us to learn from. Just consider how much it already knows about your doughnut preferences. 2. What do you do? Are you a student, developer, educator, creative professional, writer, etc.? I am a citizen of earth, the internet and all that makes up what we call today. As the Founder and CEO of , I’ve been publishing the rise via our , and integrating with next best LLM or AI API the market offers. And this past 6 months, it’s felt like there’s a new best in class option everyday. To date the via StableDiffusion (XL, 2.1, 2.0, & 1.5), Midjourney Diffusion, and Kandinsky (2.2 & 2.1), and our human writers and editors leverage ChatGPT (3, 3.5, & 4) and T-5 to provide headline and story phrasing options. Ultimately it’s the that drives how we use technology to improve stories about technology. HackerNoon tens of thousands of stories about AI HackerNoon CMS currently generates images human driven editing protocol 3. What aren’t they telling us about AI? The details that make us human can not be fully replicated. Given the training, the logic, the setting, the motivation, and every word every published on the internet, an AI can learn to guess the next word I’ll say at a high confidence level, but it can’t tell you why I’m obsessed with what Im obsessed with when I wake up. The chip manufacturing war is very real. GPU (graphics processing units) is digital international currency, and I expect it to outperform the dollar and bitcoin over the next few years. I’m excited and nervous for how much of this revolution is happening in public. There is no secret Manhattan Project today, leading us to a new level of doom; we’re doing it together. When one tech giant is behind, they simply open source the next model to catch up. Then any independent project can make small tweaks on that level up. And so I anticipate small projects by individuals can compete on a performance basis (but not on a GPU level) with large investments by large corporations. Think of this way, if LLMs were classified like new drugs and medicines, the maker would have something like a 7 year window to exclusively sell it. What they’re not telling you about AI boom of the early 2020s is that it is in fact driven by transparency of documentation. 4. What do you think the world will be like when AI reaches its full potential? The world will never reach it’s full potential. Society’s full potential, just as inbox zero, is a myth perpetuated by rounding errors. Did AI write that? No, but what’s the difference? We’ve integrated a number of different AI writing detection applications tools within the HackerNoon CMS, and they certainly help identify patterns, but they all function in confidence levels. Like what is the difference between 83% human written / 17% machine written and 84% human written / 16% machine written? The dangers as AI scales is being mistaken for a human. From the AI’s perspective, I don’t think it would see becoming human as its full potential, but my perspective may be warped by AI Scarlett Johansson (Samatha) dumping real live human Joaquin Phoenix (Theodore) towards the end of Spike Jonze’s “Her.” 5. Do you have any final thoughts about how the AI Boom of early 2020s will be remembered? The leading technology of the day usually wows the laymen on an input to output level. Generative AI can wow you. From a single word or phrase, it can create a new image at a professional or semi-professional level. It can make a solid college level lesson plan, create new tasty recipes, and even optimize your calendar. Ultimately, I believe and hope AI can assist us. Broaden our will so to speak. But , an AI version of me would very much say the same thing. When we look back on the AI boom of the early 2020s, we’ll say, if has Skynet has taken over of course machines could make new things too. Would you like to take a stab at answering some of these questions? The link for , just start writing! ”What They Don’t Say About AI” Story Template is HERE , Founder & CEO of . David Smooke HackerNoon . Publish a story with us

The 2020s AI Boom is Driven by Transparency of Documentation 

2020 এর AI বুম ডকুমেন্টেশনের স্বচ্ছতার দ্বারা চালিত

Der KI-Boom der 2020er Jahre wird durch die Transparenz der Dokumentation vorangetrieben

El auge de la IA de la década de 2020 está impulsado por la transparencia de la documentación

Le boom de l’IA des années 2020 est motivé par la transparence de la documentation

2020 का एआई बूम दस्तावेज़ीकरण की पारदर्शिता से प्रेरित है

2020 年代の AI ブームはドキュメントの透明性によって推進される

2020년대 AI 붐은 문서의 투명성이 주도

O boom da IA na década de 2020 é impulsionado pela transparência da documentação

Бум искусственного интеллекта 2020-х годов обусловлен прозрачностью документации

2020'deki Yapay Zeka Patlaması Belgelerin Şeffaflığından Kaynaklanıyor

Sự bùng nổ AI của những năm 2020 được thúc đẩy bởi tính minh bạch của tài liệu

文档透明度推动 2020 年代 AI 繁荣

Learn More at EditingProtocol.com, where the HackerNoon CMS powers the most up-to-date technical documentation about The Editing Protocol , the independent technology publishing company, has released to the public, a set of rules and guidelines that can be used by humans and machines to determine whether a story is worth publishing, how to specifically improve the story’s content, and how to distribute the story with more reach and relevance. HackerNoon The Editing Protocol What is The Editing Protocol? One of the world leaders in online publishing, HackerNoon is home to over 45,000 published contributors. As a startup, the company was tasked with reviewing thousands of submissions a month with just a small team of editors. To do that, HackerNoon created a set of preliminary checks, rules, and quality guidelines that determine whether or not a story can be published, or should be rejected. This document was dubbed The Editing Protocol and can be read by humans to guide publishing at scale. “In publishing a hundred thousand stories, we’ve learned best practices for how to improve and distribute professional technology content on the internet,” said . HackerNoon Creator and CEO David Smooke “I’m excited to open up this technical documentation for feedback, and continue integrating the most cutting-edge technologies into the Editing Protocol.” To elevate the quality of published content, and optimize the user experience for contributors and readers alike, the editing protocol incorporates several inventive technologies, such as: An AI-powered (GPT-4) headline generator. AI image generator models like Stable Diffusion XL, Stable Diffusion v2.1, Stable Diffusion v2.0, Stable Diffusion v1.5, Midjourney Diffusion, Kandinsky 2.2, and Kandinsky 2.1. AI-generated content detection. AI text translator. AI TLDR generator. Plagiarism checker. Backlink checker bots. The protocol’s guidelines are rule-based and can be easily converted into conditional statements. Once programmed, the protocol can allow human editors to focus on the quality and improvement of publishable stories, while the system automatically informs writers of what rule they have broken, or guidelines they have missed. With The Editing Protocol, small teams can publish content at scale, using both human editors and rule-based flagging systems to provide an efficient publishing process. HackerNoon has already programmed a lot of the protocol into the custom CMS the company has built from the ground up. For instance, submissions that are below the minimum structural quality measure are automatically rejected, and writers are sent an email specifically stating how to improve the story they hope to publish before resubmitting. Furthermore, there is a section in the protocol that highlights the importance of . As one of the most important aspects of online publishing is the visual presentation of the story, HackerNoon built an into their CMS, allowing writers to create original images and help writers adhere to the protocol using the tech itself. originality score AI image generator HackerNoon believes the protocol will not only guide human editors to better vet stories, but also help developers automate the tedious workflows within the traditional editorial review process. How to Use The Editing Protocol The full protocol can be viewed via which is hosted on HackerNoon’s custom CMS builder. https://editingprotocol.com/ With The Editing Protocol, HackerNoon aims to help make the internet better by getting rid of the noise—low-quality self-published articles, spammy content riddled with even spammier links, and a horde of other bad SEO practices that prioritize clicks at the expense of quality. As the Internet changes with , it is likely that The Editing Protocol, too, will change. As such, whenever rules or processes are added, removed, or updated, users can find them on , and use that domain as the ground truth. the emergence of Web3 technologies editingprotocol.com

HackerNoon Releases “The Editing Protocol” - Technical Documentation for Digital Publishing at Scale

হ্যাকারনুন "দ্য এডিটিং প্রোটোকল" রিলিজ করেছে - স্কেলে ডিজিটাল পাবলিশিংয়ের জন্য প্রযুক্তিগত ডকুমেন্টেশন

HackerNoon veröffentlicht „The Editing Protocol“ – technische Dokumentation für digitales Publishing im großen Maßstab

HackerNoon, 대규모 디지털 출판을 위한 기술 문서인 "편집 프로토콜" 출시

HackerNoon выпускает «Протокол редактирования» — техническую документацию для масштабной цифровой публикации

HackerNoon "Düzenleme Protokolü"nü Yayınladı - Büyük Ölçekte Dijital Yayıncılık için Teknik Dokümantasyon

DOE v. Github (original complaint) Court Filing, retrieved on November 3, 2022 is part of . You can jump to any part in this filing here. This is the table of links with all parts. HackerNoon’s Legal PDF Series Case Number: 3:22-cv-06823-KAW Plaintiff: DOE Defendant: Github Presiding Judge: Jon S Tigar Referring Judge: Kandis A Westmore Filing Date: November 3, 2022 Location: US District Court for the Northern District of California TABLE OF CONTENTS INTRODUCTION I. OVERVIEW: A BRAVE NEW WORLD OF SOFTWARE PIRACY II. JURISDICTION AND VENUE III. INTRADISTRICT ASSIGNMENT IV. PARTIES Plaintiffs Defendants V. AGENTS AND CO-CONSPIRATORS VI. CLASS ALLEGATIONS A. Class Definitions B. Numerosity C. Typicality D. Commonality & Predominance E. Adequacy F. Other Class Considerations VII. FACTUAL ALLEGATIONS A. Introduction B. Codex Outputs Copyrighted Materials Without Following the Terms of the Applicable Licenses C. Copilot Outputs Copyrighted Materials Without Following the Terms of the Applicable Licenses D. Codex and Copilot Were Trained on Copyrighted Materials Offered Under Licenses E. Copilot Was Launched Despite Its Propensity for Producing Unlawful Outputs F. Open-Source Licenses Began to Appear in the Early 1990s G. Microsoft Has a History of Flouting Open-Source License Requirements H. GitHub Was Designed to Cater to Open-Source Projects I. OpenAI Is Intertwined with Microsoft and GitHub J. Conclusion of Factual Allegations VIII. CLAIMS FOR RELIEF COUNT I COUNT II COUNT III COUNT IV COUNT V COUNT VI COUNT VII COUNT VIII COUNT IX COUNT X COUNT XI COUNT XII IX. DEMAND FOR JUDGMENT X. JURY TRIAL DEMANDED About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 5, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

DOE v. Github: Original Complaint Pertaining to Copyright Infringement, Open Source Licenses & More

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023, is part of . You can jump to any part in this filing . This is part 13 of 13. HackerNoon’s Legal PDF Series here III. CONCLUSION For these reasons, OpenAI’s motion to dismiss the claims for violations of the DMCA, intentional and negligent interference with prospective economic relations, unjust enrichment, unfair competition, and negligence in the FAC should be granted. About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI's Reply to the Amended Complaint: What Comes Next?

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023 is part of . You can jump to any part in this filing . This is part 12 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT G. Plaintiffs Fail to State a Claim for Negligence Plaintiffs have not adequately pled that OpenAI owe them any duty of care, as they would need to in order to state a claim for negligence. In their Opposition, Plaintiffs contend that OpenAI’s duty arises from Plaintiffs’ open source licenses and their unspecified “special relationship.” (Opp. at 24-25.) Neither is true under the circumstances here. Plaintiffs’ argument that “a duty may arise out of a contract” (id. at 24) ignores subsequent California case law that imposed an independent-duty requirement for tort claims predicated on a breach of contract. Plaintiffs’ cited authority is inapplicable. Plaintiff cites 6 Witkin, Summary of Cal. Law (11th ed. 2017) Torts, § 1201, which relies on Eads v. Marks, 39 Cal. 2d 807, 811-12 (1952) (where the California Supreme Court recognized a tort cause of action based on negligent failure to perform contractual duties), but recent decisions raise serious doubts about the viability of Eads. See Peregrine Pharms., Inc. v. Clinical Supplies Mgmt., Inc., No. SACV 12-1608 JGB (ANx), 2015 WL 13309286, at *8 n.11 (C.D. Cal. June 22, 2015) (observing that “the continued relevance of Eads is somewhat questionable”); Britz Fertilizers, Inc. v. Bayer Corp., No. 1:07-cv-00846-OWW-SMS, 2008 WL 341628, at *11 (E.D. Cal. Feb. 5, 2008) (determining that “Eads has been refuted by later California case law that establishes the independent duty requirement”). In Britz, the court highlighted three later-decided California Supreme Court cases that held “conduct amounting to a breach of contract only becomes tortious when it also violates a duty independent of the contract arising from principles of tort law.” See 2008 WL 13309286, at *7-9 (citing Aas v. Super. Ct., 24 Cal. 4th 627 (2000); Elrich v. Menezes, 21 Cal. 4th 543 (1999); and Freeman & Mills, Inc. v. Belcher Oil Co., 11 Cal. 4th 85 (1995)). Likewise, Plaintiffs cite Buxbom v. Smith, 23 Cal. 2d 535 (1944), but there the court did not hold that a breach of contract could support a finding of duty. See Buxbom at 548. Because Plaintiffs did not allege an independent duty distinct from any obligations under the open source licenses, they fail to state a claim for negligence. (See Opp. at 24.) Further, Plaintiffs have neither cited authority for finding a special relationship under these circumstances, nor plausibly alleged that the balance of the six factors they cite in their brief supports such a relationship. (See id. at 24-25 (citing Vera Mona, LLC v. Dynasty Grp. USA LLC, No. EDCV 20-2615 JGB (KKx), 2021 WL 3623297, at *4 (C.D. Cal. Apr. 15, 2021) (courts look to six factors to determine whether a special relationship exists: “(1) the extent to which the transaction was intended to affect the plaintiff, (2) the foreseeability of harm to the plaintiff, (3) the degree of certainty that the plaintiff suffered injury, (4) the closeness of the connection between the defendant’s conduct and the injury suffered, (5) the moral blame attached to the defendant's conduct and (6) the policy of preventing future harm”)).) These factors do not support any “special relationship” based on OpenAI’s alleged use of Plaintiffs’ code for training AI. The open source nature of Plaintiffs’ code makes it unlikely that this use was intended to affect Plaintiffs, or that it was foreseeable and certain that Plaintiffs would suffer injury. Indeed, the purpose of open source is to make code publicly accessible for anyone to use and modify, as well as foster collaboration. Moreover, Plaintiffs have not cited any authority finding that using publicly available data to train and develop generative AI products is morally wrong or violates a public policy preventing such use. (See Opp. at 25.) Plaintiffs primarily argue that “to the extent Copilot generates a duty because it hosts Plaintiffs’ open source code,” that would also extend to OpenAI because Copilot is alleged to be a joint venture between GitHub and OpenAI. (Opp. at 25.) But Plaintiffs cite no cases supporting their argument that entering into a business relationship with a separate entity that hosts publicly available code somehow creates a duty to Plaintiffs. The three cases that Plaintiffs do cite are inapposite. In Witriol v. LexisNexis Group, the court ruled that plaintiffs adequately pled the duty element of negligence by alleging that “[a]s custodians of the Representative Plaintiff’s and the Class Members’ personal and confidential information,” defendants owed a duty to prevent unauthorized access to private and sensitive data. No. 05-CV-02392 MJJ, 2006 WL 4725713, at *1, 8 (N.D. Cal. Feb. 10, 2006) (emphasis added). Similarly, Bass v. Facebook, Inc. held that Facebook owed its users a duty of care because “some of the information [published on Facebook] was private.” 394 F. Supp. 3d 1024, 1039 (N.D. Cal. 2019). Neither case supports finding a duty where an entity merely “possess[es]” publicly available data of individuals with whom they lack any relationship. (See Opp. at 25.) Likewise, in In re Sony Gaming Networks & Customer Data Security Breach Litigation, the court found that defendant owed a “legal duty to safeguard consumer’s confidential information entrusted to a commercial entity.” 996 F. Supp. 2d at 955, 966 (S.D. Cal. 2014) (emphasis added). That is not the case here. Because Plaintiffs’ code is available to the public, no similar duty exists. This claim should be dismissed. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI Court Case: Do the Plaintiffs Have a Claim for Negligence?

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023 is part of . You can jump to any part in this filing . This is part 11 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT F. Plaintiffs Fail to State an Unfair Competition Claim Plaintiffs have not stated a UCL claim under any prong.[2] Plaintiffs do not dispute that the FAC failed to establish an inadequate remedy at law sufficient to state a UCL claim. (See MTD at 21; Opp. at 23.) This alone warrants dismissal of their UCL claim. See Sharma, 524 F. Supp. 3d at 907 (concluding that, under Sonner v. Premier Nutrition Corp., 971 F.3d 834, 844 (9th Cir. 2020), plaintiffs “must establish that they lack an adequate remedy at law before securing equitable restitution” under the UCL) (cleaned up); Lou v. Am. Honda Motor Co., Inc., No. 16- CV-04384-JST, 2022 WL 18539358, at *2 (N.D. Cal. Aug. 26, 2022) (finding plaintiff “concedes” argument when plaintiff “offers no response to [defendant’s] argument that he failed to plead an inadequate remedy at law”). Moreover, Plaintiffs’ argument that they were “denied their license, permission, and recognition” does not establish economic injury sufficient for UCL standing. (Opp. at 23.) Standing under the UCL is “substantially narrower than federal standing under article III” and has “more stringent” requirements. Kwikset Corp. v. Super. Ct., 51 Cal. 4th 310, 324 (2011). Intangible losses are not enough. In re Yahoo! Inc. Customer Data Sec. Breach Litig., No. 16- MD-02752-LHK, 2017 WL 3727318, at *22 (N.D. Cal. Aug. 30, 2017) (Plaintiffs’ “imminent risk of future costs as a result of the [d]ata [b]reaches” fail to “show that [p]laintiffs have specifically lost money or property as a result of [d]efendants’ misconduct.”) (cleaned up). Labeling alleged intangible losses as “economic benefits” does not create standing; “generating market share for programs, increasing national or international reputation…, and deriving value from improvements to software based on suggestions by end-users" do not constitute “lost money or property” under the UCL. (See FAC ¶¶281, 243.) Plaintiffs’ cited cases do not show otherwise. Jacobsen and Arifex Software did not involve UCL claims. Jacobsen, 535 F.3d at 1377 (copyright infringement); Artifex Software, Inc. v. Hancom, Inc., No. 16-CV-06982-JSC, 2017 WL 4005508, at *2 (N.D. Cal. Sept. 12, 2017) (breach of contract). Rather, Jacobsen and Arifex Software merely noted that the “lack of money changing hands in open source licensing should not be presumed to mean there is no economic consideration” in an open source agreement. Jacobsen, 535 F.3d at 1379; Arifex, 2017 WL 4005508, at *4 (citing Jacobsen). Moreover, even if these purported “benefits” could qualify as economic injuries, Plaintiffs have not alleged that OpenAI’s conduct caused decreased “market share[s] for [Plaintiffs’ code],” decline in Plaintiffs’ “national or international reputation,” or any loss in value of Plaintiffs’ code. (See FAC ¶¶ 243, 281.) Plaintiffs’ UCL claim under the unlawful, unfair, and fraudulent prongs fails for additional reasons. For the “unlawful” prong, Plaintiffs do not dispute that their derivative UCL claim fails if the predicate DMCA and state law claims fail. (See MTD at 21; Opp. at 22-23.) For the “unfair” prong, Plaintiffs’ conclusory assertion that OpenAI’s use of Plaintiffs’ code for training is an “immoral, unethical, oppressive, or unscrupulous business practice[]” that “outweighs the utility of the practice” (Opp. at 23.) is deficient. See Doe v. CVS Pharmacy, Inc., 982 F.3d 1204, 1215 (9th Cir. 2020) (affirming dismissal of unfair claim where plaintiffs “assert in a conclusory fashion that [defendant’s] conduct outweighs any justification, motive or reason therefor, but they do not allege how that is so”) (cleaned up); Nazemi v. Specialized Loan Servicing, LLC, No. 2:22-cv-05006-MCS-PVC, 2022 WL 17220707, at *4 (C.D. Cal. Oct. 31, 2022) (“Plaintiff’s bare conclusory recitation of one of the UCL’s legal standards does not clarify how Defendants’ conduct satisfies” the tests for the unfair prong.) (cleaned up). For the “fraudulent” prong, Plaintiffs do not dispute that they must demonstrate actual reliance and satisfy the heightened pleading requirements under Federal Rule of Civil Procedure 9(b) on the purported deception to state a UCL claim. (See MTD at 22; Opp. at 23.) As Plaintiffs’ own cited case shows, plaintiffs must plead “actual reliance on the alleged fraudulent conduct” and plead “with specificity to meet Rule 9(b)’s strictures.” Garcia v. Sony Comput. Ent. Am., LLC, 859 F. Supp. 2d 1056, 1063, 1066 (N.D. Cal. 2012). Plaintiffs have not done so here. See Opperman v. Path, Inc., 84 F. Supp. 3d 962, 974 (N.D. Cal. 2015) (claims sounding in fraud must contain “an account of the time, place, and specific content of the false representations as well as the identities of the parties to the misrepresentations”). Their UCL claim should be dismissed. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

The OpenAI Case: What the Company Has to Say Against the Plaintiffs' Claims

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023, is part of . You can jump to any part in this filing . This is part 10 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT E. Plaintiffs Fail to State a Claim for Unjust Enrichment Plaintiffs’ unjust enrichment claim fails because they have not alleged the absence of any applicable and enforceable contract provisions, even in the alternative, as required for quasi-contract claim. Plaintiffs do not deny that an unjust enrichment claim cannot proceed “when the parties have a valid contract regarding the same subject matter.” (See MTD at 19.) Plaintiffs allege that the open source licenses are a contract between them and OpenAI and that OpenAI breached those licenses. (FAC ¶¶ 217, 222-25.) Although Plaintiffs now argue in their Opposition that “the [open source] licenses do not specifically address the use of code as training data or in connection with generative AI products sold to the public” (Opp. at 21), nowhere in the FAC do Plaintiffs allege that the open source licenses are inapplicable. See In re Bang Energy Drink Mktg. Litig., No. 18-CV-05758-JST, 2020 WL 4458916, at *10 (N.D. Cal. Feb. 6, 2020) (dismissing unjust enrichment claim where “[p]laintiffs have not alternatively alleged the absence of [applicable and enforceable] provisions”). Moreover, Plaintiffs do not sufficiently allege that OpenAI unjustly retained a benefit at their expense. (See MTD at 20.) “Restitution [under unjust enrichment] is not mandated merely because one person has realized a gain at another’s expense,” but rather, “the obligation arises when the enrichment obtained lacks any adequate legal basis and thus cannot conscientiously be retained.” Russell v. Walmart, Inc., No. 22-CV-02813-JST, 2023 WL 4341460, at *2 (N.D. Cal. July 5, 2023) (cleaned up). Restitution requires “that a defendant has been unjustly conferred a benefit,” and “[a]bsent qualifying mistake, fraud, coercion, or request by [defendant], there is no injustice.” Id. (cleaned up). Plaintiffs fail to point to any such qualifying conduct by OpenAI to render retention of any benefit unjust. See id. (dismissing unjust enrichment claim based on allegations that customers using Walmart’s self-checkout conferred a benefit on Walmart by providing uncompensated labor because plaintiff failed to allege qualifying mistake, fraud, coercion, or request by Walmart). This claim should be dismissed. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI: Did the Plaintiffs Fail to State a Claim for Unjust Enrichment?

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023 is part of . You can jump to any part in this filing . This is part 9 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT D. Plaintiffs’ Intentional and Negligent Interference with Prospective Economic Relations Claims Fail The Opposition does not point to factual allegations to satisfy a single element of Plaintiffs’ claims for intentional or negligent interference with prospective economic advantage. Plaintiffs contend that they have satisfied this element because (i) the FAC refers to “open-source licensing relationships and communities,” (Opp. at 19), and (ii) courts have recognized that there are “economic benefits[] to the creation and distribution of copyrighted works under public licenses.” Jacobsen v. Katzer, 535 F.3d 1373, 1379 (Fed. Cir. 2008). No Economic Relationship with the Probability of Future Economic Benefit. While licenses may sometimes confer benefits, Plaintiffs’ cited authority does not involve a tortious interference claim and does not stand for the proposition that referring to open source licenses alone is sufficient to satisfy this element at the pleading stage. As stated in OpenAI’s Motion, Plaintiffs must plead specific economic relationships between Plaintiffs and third parties that have the probability of future economic benefit, not merely the hope of future transactions. (See MTD at 15-16.) Plaintiffs have failed to do so and have not cited a single case that states otherwise. Plaintiffs’ conclusory argument that Defendants were aware of “these relationships, including those of open-source communities” (Opp. at 19.) fails to show that they adequately alleged OpenAI’s knowledge of an economic relationship between Plaintiffs and a third party. No Knowledge of an Economic Relationship with Third Parties. Further, while Plaintiffs argue they need not name the third party (id. at 20), the Ramona court made clear that the third party must be “identified in some manner.” Ramona Manor Convalescent Hosp. v. Care Enters., 177 Cal. App. 3d 1120, 1133 (1986). As many courts have acknowledged, the allegations “must do more than conclusorily allege the existence of business relationships with which the defendant interfered.” Logistick, Inc. v. AB Airbags, Inc., 543 F. Supp. 3d 881, 887 (S.D. Cal. 2021); see also Damabeh v. 7-Eleven, Inc., No. 5:12-cv-1739-LHK, 2013 WL 1915867, at *10 (N.D. Cal. May 8, 2013) (a plaintiff “must identify with particularity the relationships or opportunities with which defendant is alleged to have interfered”). Plaintiffs’ allegations do not identify with particularity the relationships with which OpenAI allegedly interfered. Plaintiffs’ conclusory statements that the FAC “alleges actual disruption of Plaintiffs’ relationship with their users” and “does adequately allege economic harm due to interference with Plaintiffs’ open source code community” are insufficient to defeat OpenAI’s Motion. (Opp. at 20.) No Actual Disruption and Economic Harm. These blanket assertions, without any factual support, do not provide the requisite showing that OpenAI’s conduct actually disrupted Plaintiffs’ relationships with third parties or resulted in economic harm. See Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555 n.3 (2007) (“Rule 8(a)(2) still requires a ‘showing,’ rather than a blanket assertion.”). For example, Plaintiffs do not allege in what way their future relationships with specific third parties were disrupted. Further, Plaintiffs’ statement that “[t]he fact that the open-source software was offered for free does not mean there was no consideration owed under the open-source license” does not demonstrate that Plaintiffs have suffered economic harm. (Opp. at 20.) As stated in OpenAI’s Motion, Plaintiffs have not identified any prospective contracts, job offers, or research assignments that Plaintiffs allegedly lost due to OpenAI’s purported interference, or the alleged monetary and reputational harm Plaintiffs experienced. (MTD at 18.) Plaintiffs have no response to their failure to plead “conduct that was wrongful by some legal measure other than the act of interference itself.” (MTD at 18.) No Wrongful Conduct. Plaintiffs refer to OpenAI’s alleged failure to attach licenses when emitting code (see Opp. at 20), but that conduct is not separate and distinct from the act of interference itself as alleged in the FAC. (See FAC ¶ 262 (alleging that OpenAI’s operation of Codex and reproduction of code without correct licenses “depriv[ed] Plaintiffs of the economic benefits of open-source distribution”).) Plaintiffs’ argument fails on its face. Plaintiffs have failed to demonstrate that the FAC sufficiently pleads “acts on the part of the defendant designed to disrupt the relationship” to satisfy the intent element of an intentional interference claim. Korea Supply Co. v. Lockheed Martin Corp., 29 Cal. 4th 1134, 1154 (2003). No Intentional Acts. Rather, Plaintiffs conclude that OpenAI had the requisite intent based on its alleged “failure to attach open-source licenses when emitting code.” (Opp. at 20.) Those allegations, however, are insufficient because they do not allege facts that show that OpenAI’s conduct was designed to disrupt Plaintiffs’ relationships with unknown third parties. See name.space, Inc. v. Internet Corp. for Assigned Names & Numbers, No. CV 12-8676 PA (PLAx), 2013 WL 2151478, at *8 (C.D. Cal. Mar. 4, 2013) (dismissing claim where the complaint did not identify “any intentional actions undertaken by [defendant] designed to induce breach of Plaintiff’s contracts with its clients . . . . ” (cleaned up). Plaintiffs have failed to mention, let alone identify, sufficient factual allegations demonstrating that OpenAI owed Plaintiffs a purported duty of care, which is fatal to Plaintiffs’ negligent interference claim. No Duty of Care. See Stolz v. Wong Commc’ns Ltd. P’ship, 25 Cal. App. 4th 1811, 1825 (1994) (A negligent interference claim “arises only when the defendant owes the plaintiff a duty of care.”). Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI: Company States That Some of the Plaintiffs' Claims Fail

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023 is part of . You can jump to any part in this filing . This is part 8 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT C. Plaintiffs Fail to Plead a DMCA Claim Under Section 1202(b) 2. Plaintiffs fail to allege removal from identical copies. In response to OpenAI’s authority from the Ninth Circuit that they must allege removal of CMI from identical copies (MTD at 12-14), Plaintiffs cite only to out-of-circuit cases that do not support their position. (See Opp. at 14-15.) See Bounce Exch., Inc. v. Zeus Enter., Ltd., No. 15cv3268 (DLC), 2015 WL 8579023 (S.D.N.Y. Dec. 9, 2015) (addressing whether terms appearing in source code were CMI, not whether removal from non-identical copies falls under § 1202); ICONICS, Inc. v. Massaro, 192 F. Supp. 3d 254 (D. Mass. 2016) (addressing whether source code headers were CMI, not whether copy must be identical). Plaintiffs next argue “[t]hat the emissions were not exact or identical is of no moment when the allegations make clear the Defendants copied Plaintiffs’ code in order to train Codex and Copilot.” (Opp. at 15.) Plaintiffs appear to be arguing that the cases about identical copies are really about whether an inference of copying can be made, and that they do not apply where Plaintiffs have alleged copying of their original work in its original form such that an inference is not required. (Id.) But these cases aren’t about whether an inference is required—they are about whether excerpting or modifying a work constitutes removal of CMI at all, and hold that it does not. (MTD at 12-13.) Finally, Plaintiffs selectively quote from the original summary judgment opinion in FrostTsujiArchitects v. Highway Inn, Inc. (Opp. at 15), leaving out the vital remainder of the quoted sentence: “‘Virtually identical’ plans could have been created by redrawing Frost-Tsuji’s plans and not including Frost-Tsuji’s copyright management information, but that would not involve removal or alteration of copyright management information from Frost-Tsuji’s original work.” No. 13-00496 SOM/BMK, 2014 WL 5798282, at *5 (D. Haw. Nov. 7, 2014) (language omitted from Opp. bolded). The Frost-Tsuji court did not “[make] plain” that copies could be non-identical, as Plaintiffs contend. (Opp. at 15.) Rather, the court found the opposite—that creating virtually identical plans without CMI would not constitute removal from an original work. See Frost-Tsuji, 2014 WL 5798282 at *5. Plaintiffs also misread Dolls Kill, Inc. v. Zoetop Bus. Co., No. 22-cv-01463, 2022 WL 16961477 (C.D. Cal. Aug. 25, 2022), as demonstrating “the lack of identicality can only undermine the inference there was copying when no such allegations are made.” (Opp. at 15.) Dolls Kill says no such thing, and in fact suggests that the DMCA should not be implicated at all here, where Codex or Copilot is generating nonidentical copies. 2022 WL 16961477, at *3 (“Re-creating another party’s work may be unlawful, but it does not necessarily implicate the DMCA because copying a work does not require the removal or alteration of CMI.”). Because Plaintiffs have failed to plead removal of CMI and affirmatively allege nonidentical outputs, their § 1202 claim should be dismissed with prejudice. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI Argues That the Plaintiffs' 1202 Claim Should Be Dismissed

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023 is part of . You can jump to any part in this filing . This is part 7 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT C. Plaintiffs Fail to Plead a DMCA Claim Under Section 1202(b) 1. Plaintiffs have not pled removal of CMI. Plaintiffs argue that their DMCA claim cannot be dismissed because they offer more detail in the FAC than their original Complaint. (Opp. at 13.) But Plaintiffs’ newly pled facts confirm the implausibility of Plaintiffs’ DMCA claim. Plaintiffs do not dispute that, in order to get Copilot to output modified portions of Plaintiffs’ source code, Plaintiffs first had to input very specific and substantial portions of Plaintiffs’ source code into Copilot. (MTD at 11-12 (discussing example of Copilot suggesting five lines of modified code only after Plaintiffs input eight lines of verbatim code).) Plaintiffs offer no explanation for how Copilot’s suggestion of a modified snippet of Plaintiffs’ code could constitute OpenAI’s removal of CMI from Plaintiffs’ work, when Plaintiffs themselves provided the majority of the work to Copilot as the input without any CMI. In other words, if there was any CMI on the code in Plaintiffs’ examples in the first place, Plaintiffs (not OpenAI) removed it, because in those examples Plaintiffs provided the majority of the work to Copilot without any CMI. By Plaintiffs’ logic, Plaintiffs could input into Copilot 99 out of 100 lines of a work’s code without CMI, and if Copilot suggests an output thatresembles the work’s 100th line of code without including CMI, that output of a single line of code would violate the DMCA, despite the fact that it was Plaintiffs who provided the first 99 lines of the work without CMI. Plaintiffs provide no authority to support their argument that OpenAI must reproduce CMI in such circumstances when Plaintiffs are responsible for providing significant portions of the work as input without CMI. Plaintiffs have thus failed to plead removal of CMI. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI: Plaintiffs Have Not Pled Removal of CMI - What It All Means

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023, is part of . You can jump to any part in this filing . This is part 6 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT C. Plaintiffs Fail to Plead a DMCA Claim Under Section 1202(b) Plaintiffs spend much of the DMCA section of their Opposition (Opp. at 11-16) addressing arguments that OpenAI did not make in its motion. (See MTD at 10-14.) Defendants are not relitigating issues the Court has ruled on. Rather, Defendants’ DMCA arguments are limited to two issues pertaining to newly pled facts in the FAC: (1) Plaintiffs’ specific examples of “removal” are not removal of CMI from a work at all; and (2) the alleged “removals” from these added examples are not from identical copies of Plaintiffs’ works.[1] (Id.) Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI Court Case: Company Claims Plaintiffs Fail to Plead a DMCA Claim

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023, is part of You can jump to any part in this filing . This is part 5 of 13. HackerNoon’s Legal PDF Series. here II. ARGUMENT B. The Copyright Act Preempts Plaintiffs’ State Law Causes of Action Plaintiffs affirmatively assert that their California law claims are “premised on the unauthorized use of Plaintiffs’ code for training purposes by Codex and Copilot without their consent and removal of their CMI.” (Opp. at 16; see also id. at 17 (stating that “it is that wrongful use [of Plaintiffs’ code as training data] that [] underlies Plaintiffs’ California law claims”).) Plaintiffs argue that the right to use Plaintiffs’ code for training AI products places their claims beyond the reach of federal copyright law because Section 106 of the Copyright Act doesn’t cover “use.” (Id. at 17.) This is an admission that training AI models does not, in and of itself, infringe anyone’s copyrights. OpenAI agrees, but Plaintiffs’ claims are still preempted. Where a state law claim arises from a right within the general subject matter of copyright, it can still be subject to preemption, even if the precise contours of the right differ from those conferred by Section 106 of the Copyright Act. See Close v. Sotheby’s, Inc., 894 F.3d 1061, 1070-71 (9th Cir. 2018) (holding that a California law that granted rights beyond the scope of a copyright holder’s distribution right nonetheless was equivalent to the distribution right for preemption purposes); see also Symantec Corp. v. McAfee Assocs., No. C-97-20367, 1998 WL 740798, *4 (N.D. Cal. Jun. 9, 1998) (The “fact that the state-created right is ... broader ... than its federal counterpart will not save it from pre-emption.”) (cleaned up). Plaintiffs do not appear to dispute that their claims fall within the subject matter of copyright. The question, then, is whether Plaintiffs’ claims are “equivalent to rights within the general scope of copyright as specified by section 106 of the Copyright Act.” Close, 894 F.3d. at 1069 (cleaned up). Notably, a plaintiff’s claims need not be “coextensive” with or “precisely within the contemplation of” the Copyright Act to be considered “equivalent” to the rights within Section 106 of the Copyright Act, id. at 1071; rather, claims are preempted unless they are “qualitatively different from the copyright rights,” Laws v. Sony Music Ent., Inc., 448 F.3d 1134, 1143 (9th Cir. 2006) (cleaned up). In arguing that that standard is met here, Plaintiffs claim that their state-law claims are about “use” of their code as training data, and that “use” is not one of the exclusive rights of copyright. (Opp. at 16 (citing MTD Order at 10).) But Plaintiffs made essentially the same argument with respect to their original Complaint. (See Dkt. No. 66 at 11 (“Plaintiffs allege that Defendants, through their unauthorized use of Plaintiffs’ code to train Codex and Copilot, and their display of Plaintiffs’ code to others for commercial gain, violated Plaintiffs’ rights under state and common law.”).) This Court rejected Plaintiffs’ argument then, and it should do so now. Plaintiffs’ claims are fundamentally constructed on the premise that they own copyrighted code, and that that code was misappropriated by OpenAI to create unauthorized reproductions and an unauthorized derivative work. (See MTD at 8-9.) Those claims are “equivalent” to a claim based on rights within Section 106 of the Copyright Act, and Plaintiffs cannot escape that fact merely by attempting to label them as “use.” For example, in Laws, the plaintiff argued that her claim was not preempted because it was based only on unauthorized “use” of her voice, but the claim was nonetheless preempted, because the “the sole basis for her voice misappropriation claim is the unauthorized reproduction of her copyrighted vocal performance,” whether or not that reproduction infringed anyone’s copyright. Laws, 448 F.3d at 1143-44. Altera Corp. v. Clear Logic, Inc., 424 F.3d 1079 (9th Cir. 2005), the case cited by the Court in its prior order (MTD Order at 17), does not compel a different conclusion. In Altera, customers using the plaintiff’s software, which programmed semiconductor chips, agreed to the terms of a licensing agreement that limited the use of software to “the sole purpose of programming [chips] manufactured by … and sold by [the plaintiff].” 424 F.3d at 1082. In finding that plaintiff’s breach-of-contract claim was not preempted, the court noted that the underlying conduct challenged was a prohibited use of the software’s output bitstream files, rather than reproduction of the software itself. Id. at 1089. Here, Plaintiffs have not challenged OpenAI’s use of any end-product of their code. Rather, Plaintiffs’ claims rest on the supposed unauthorized creation of a derivative work (in the form of the model) and reproduction of Plaintiffs’ code (in the form of outputs from the model). Such claims are preempted. See DocMagic, Inc. v. Ellie Mae, Inc., 745 F. Supp. 2d 1119, 1148 (N.D. Cal. 2010) (concluding that, because the “unauthorized use of a software’s end-product is, however, legally and factually distinct from the unauthorized use of software itself,” defendant’s counterclaim related to the unauthorized use of copyrighted software constitutes copyright infringement). Therefore, all of Plaintiffs’ state law claims, except for their breach of contract claim, must be dismissed because they are preempted. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

OpenAI: What Does the Law Actually Say?

Open AI reply to amended complaint Court Filing Kandis A. Westmore, November 3, 2023 is part of . You can jump to any part in this filing . This is part 4 of 13. HackerNoon’s Legal PDF Series here II. ARGUMENT A. Plaintiffs Cannot Allege Any Injury-in-Fact to Establish Article III Standing 3. Plaintiffs’ hunch that their code has been output does not justify jurisdictional discovery. In addition, Plaintiffs have not shown that they are entitled to any jurisdictional discovery to cure their standing defects. (Opp. at 10.) While courts have discretion to grant or deny discovery, discovery is inappropriate where the request for discovery is "based on little more than a hunch that might yield jurisdictionally relevant facts," or "when it is clear that further discovery would not demonstrate facts sufficient to constitute a basis for jurisdiction." Gomez v. Elite Lab. Servs. Weeklys, Ltd., No. 20-CV-01805-JST, 2021 WL 4992625, at *6 (N.D. Cal. Mar. 16, 2021)). "Absent some evidence establishing a colorable basis for jurisdiction," a court must deny a plaintiff's request for jurisdictional discovery. Sharma v. Volkswagen AG, 524 F. Supp. 3d 891, 905 (N.D. Cal. 2021) (cleaned up). Plaintiffs cannot meet these minimal standards. The Opposition does not provide any "colorable basis" or arguments beyond a mere "hunch" that such discovery would establish standing for Does 3 and 4. Any request for jurisdictional discovery must be denied. Continue Reading Here. About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 3:22-cv-06823-KAW retrieved on September 2, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Storage.Courtlistener

Stories, News & Mentions about KANDJI

Kandji’s Stories on HackerNoon

Kandji's News & Mentions