Parallel Domain’s API lets Customers use Generative AI to build Synthetic Datasets

Interesting Direction

Parallel Domain’s API lets customers use generative AI to build synthetic datasets

Rebecca Bellan@rebeccabellan / 11:00 AM EDT•June 19, 2023

Parallel Domain is putting the ability to generate synthetic datasets into the hands of its customers. The San Francisco-based startup has launched a new API called Data Lab that stands on the shoulders of generative AI giants, giving machine-learning engineers control over dynamic virtual worlds to simulate any scenario imaginable. 

“All you have to do is you go to GitHub, you install the API, and then you can start writing Python code that generates datasets,” Kevin McNamara, founder and CEO of Parallel Domain, told TechCrunch.

Data Lab allows engineers to generate objects that weren’t previously available in the startup’s asset library. The API uses 3D simulation to provide a foundation upon which an engineer, through a series of simple prompts, can layer the real world in all its randomness on top. Want to train your model to drive on a highway with a cab flipped over across two lanes? Easy. Think your robotaxi should know how to identify a human dressed in an inflatable dinosaur outfit? Done.

A 5G eSIM cloud platform allows for seamless connectivity across borders for businesses wishing to scale rapidly in the global marketplace

The goal is to give autonomy, drone and robotics companies more control over and more efficiency in building large datasets so they can train their models quicker and at a deeper level.

“Iteration time now goes to essentially how fast can you, as an ML engineer, think of what you want and translate that into an API call, a set of code?” said McNamara. “There is a near infinite, unbounded level of stuff a customer could type in for a prompt, and the system just works.”

API for ChatGPT for Enterprise Customers

 More advances, direct    

OpenAI launches an API for ChatGPT, plus dedicated capacity for enterprise customers

Kyle Wiggers@kyle_l_wiggers /   inTechCrunch

To call ChatGPT, the free text-generating AI developed by San Francisco-based startup OpenAI, a hit is a massive understatement.

As of December, ChatGPT had an estimated over 100 million monthly active users. It’s attracted major media attention and spawned countless memes on social media. It’s been used to write hundreds of ebooks in Amazon’s Kindle store. And it’s credited with coauthoring at least one scientific paper.

But OpenAI, being a business — albeit a capped-profit one — had to monetize ChatGPT somehow, lest investors get antsy. It took a step toward this with the launch of a premium service, ChatGPT Plus, in February. And it made a bigger move today, introducing an API that’ll allow any business to build ChatGPT tech into their apps, websites, products and services.

Markets are shifting and business models are adapting. Learn why Revenue Lifecycle Management is a game changer to your career and the future of your company.

An API was always the plan. That’s according to Greg Brockman, the president and chairman of OpenAI (and also one of the co-founders). He spoke with me yesterday afternoon via a video call ahead of the launch of the ChatGPT API.  ... ' 

OpenAI Releases DALL-E 2 API for Developers

More from OpenAI

OpenAI Releases DALL-E 2 API for Developers

ERIC HAL SCHWARTZ on November 3, 2022 at 2:00 pm   in Voicebot.ai

OpenAI has released an API of its text-to-image engine DALL-E 2 as a public beta. Developers can now embed the synthetic media generator into apps and websites, paying the company based on the number and size of images its customers generate.

DALL-E API

The DALL-E 2 API works much like the service on OpenAI’s website, turning text into images that can then be edited. OpenAI claims DALL-E is used by more than three million people to generate more than four million images every day. OpenAI’s DALL-E image generator is already coming to Microsoft products, including the 365 Office suite, Bing, and Edge, while Shutterstock announced its own plans to incorporate DALL-E into its platform. OpenAI also pointed to smaller companies like photography startup Mixtiles and fashion design app CALA as examples of businesses that will use the DALL-E 2 API. Companies using the DALL-E 2 API pay for each image generated, with the price rising by the image size. Each 256×256 image costs $0.016, rising to $0.018 for 512×512 images, and $0.02 for 1024×1024 pictures.

“DALL·E excels at following natural language descriptions so users can plainly describe what they want to see. As our research evolves, we will continue to bring the state of the art into the API, including advances in image quality, latency, scalability, and usability,” OpenAI explained in its announcement. “Mixtiles uses the DALL·E API to create and frame emotionally resonating artwork, by guiding users through a creative process that captures childhood memories, dream destinations, and more. We’re excited to see what our customers will do with DALL·E and what creative ideas they’ll come up with.”  ...  ' 

Reimplementing Software Interfaces is Fair Use

 Somewhat surprising,  and Beyond the US?   Good overview of the technical implications.

Reimplementing Software Interfaces Is Fair Use

By Pamela Samuelson  Communications of the ACM, July 2021, Vol. 64 No. 7, Pages 24-26

10.1145/3466607

A long-standing, generally accepted norm in the computing field distinguishes between software interfaces and implementations: Programmers should have to write their own implementing code, but they should be free to reimplement other developers' program interfaces. This norm, of which Sun Microsystems, the developer of Java, was once the software industry's foremost proponent, is now the law of the land in the U.S. after the Supreme Court's decision in Google Inc. v. Oracle America, Inc., which overturned a lower court ruling that reimplementing an interface infringed copyright.

The Supreme Court took Google's appeal on two issues. One was whether program interfaces are protectable by copyright law. The Supreme Court declined to decide that issue, even though many amicus curiae (friend of the court) briefs filed by software developers, organizations such as the Electronic Frontier Foundation, the Center for Democracy & Technology, and the Computer & Communications Industry Association, as well as numerous intellectual property scholars, supported Google's argument that program interfaces are uncopyrightable.

A second issue was whether Google's reimplementation of 11,500 declarations from 37 Java Application Program Interface (API) packages in its Android smartphone platform was fair use or infringement. Although a jury rendered a verdict in favor of Google's fair use defense after a two-week trial, the Court of Appeals for the Federal Circuit (CAFC) overturned this verdict. The CAFC concluded that no reasonable jury could have found Google's appropriation of that many lines of computer code was fair use. This is the ruling that Supreme Court's decision reversed.

The penultimate sentence in Justice Breyer's opinion for the 6-2 majority succinctly states the Court's conclusion: "where Google reimplemented a user interface, taking only what was needed to allow users [that is, programmers] to put their accrued talents to work in a new and transformative program, Google's copying of the Sun Java API was a fair use of that material as a matter of law."

After explaining Oracle's claims against Google, this column reviews the Court's reasons for rejecting Oracle's arguments on the fair use issue.  ... ' 

API Security

Pointed out to me recently.   Have not been involved in API security, seems there are useful tips here.

Tips To Strengthen API Security  By Bill Doerrfeld in DevOps

If you haven’t noticed, digital organizations are building more and more APIs. ProgrammableWeb tracks more than 23,000 public web APIs at the time of writing, and the API market is estimated to be worth $5.1 billion by 2023. Building with APIs increases internal interoperability, reduces development time and can extend product functionality tremendously. In short, the value of APIs is rising. However, opening up with APIs brings security caveats that, if not addressed, could result in serious breaches that negate these benefits.  .... 

 ... APIs have been called “the next frontier in cybercrime.” Rightly so, as API breaches continue to pop up nearly every day. Take the recent API vulnerabilities at Cisco Systems, Shopify, Facebook, U.S. presidential campaign apps, and GCP as evidence. The most infamous was likely the Equifax breach—not enforcing formats on incoming API calls resulted in a massive data breach, which cost the company a $700 million lawsuit.  ... " 

Google releases API to train smaller, faster AI models

Interesting development, note small and faster.  I assume to make them useful for edge devices..  Also with the ability to retrain quickly.

Google releases API to train smaller, faster AI models   Kyle Wiggers  in VentureBeat

Google today released https://blog.tensorflow.org/2020/04/quantization-aware-training-with-tensorflow-model-optimization-toolkit.html Quantization Aware Training (QAT) API, which enables developers to train and deploy models with the performance benefits of quantization — the process of mapping input values from a large set to output values in a smaller set — while retaining close to their original accuracy. The goal is to support the development of smaller, faster, and more efficient machine learning models well-suited to run on off-the-shelf machines, such as those in medium- and small-business environments where computation resources are at a premium.

Often, the process of going from a higher to lower precision is noisy. That’s because quantization squeezes a small range of floating-point values into a fixed number of information buckets, leading to information loss similar to rounding errors when fractional values are represented as integers. (For example, all values in range [2.0, 2.3] might be represented in a single bucket.) Problematically, when the lossy numbers are used in several computations, the losses accumulate and need to be rescaled for the next computation. .... "

Capturing the Value of APIs

Good general discussion of the architecture.   Good choices can make API possible and an asset.   In McKinsey.

What it really takes to capture the value of APIs

By Keerthi Iyengar, Somesh Khanna, Srinivas Ramadath, and Daniel Stephens

APIs are the connective tissue in today’s ecosystems. For companies who know how to implement them, they can cut costs, improve efficiency, and help the bottom line.

Application programming interfaces (APIs) were once largely limited to technical domains but have now become a significant engine of business growth. As the connective tissue linking ecosystems of technologies and organizations, APIs allow businesses to monetize data, forge profitable partnerships, and open new pathways for innovation and growth.

Early adopters across industries are already using APIs to create new products and channels and improve operational efficiency. Within the automotive industry, for instance, APIs are used to embed efficiency data, driving statistics, route information and real-time alerts into dashboards. Some retailers are using APIs to set up multi-brand shopping platforms, track inventory, and help consumers locate stores. And a handful of banks are partnering with fintechs and retailers, among others, to develop APIs that help customers integrate banking data into bookkeeping and investment software, and provide faster internal access to a range of account information. .... "

Watson Analyzing Social Tone

Instructive Watson Example, via O'Reilly

How can I use IBM Watson to analyze the tone of social media posts?
Learn how to use IBM Watson's APIs and natural language understanding to analyze the tone of social media posts like tweets.  Video.

By Matt Coatney  ... " 

APIs for Data Scientists

Good piece here,  the architecture is key, and well written APIs can make that happen.

How will APIs change the role of data scientists?  
By Gregor Herrmann on Hidden Insights   in SAS.com 

Application programming interfaces, otherwise known as APIs, are set to radically change the face of analytics. These little bits of code are working behind the scenes, hidden treasures that allow two pieces of code to talk to each other. This means that developers can borrow functionality from other apps. For example, Facebook’s API means that other apps can use its standard sign-in process. Google Maps’ API allows other apps to draw on its maps and location-finding technology, and not have to develop their own.

The potential impact of APIs in analytics is huge. From gathering data from new sources, through modelling and getting the results out to consumers, to evaluation, they can be beneficial to various steps across the whole analytical lifecycle. And this, in turn, is changing the role of data scientists, including citizen data scientists. .... " 

API Economy

Expanding business models

2017 Is Quickly Becoming The Year Of The API Economy

This year more CIOs will have their bonuses tied to how many new business models they help create with existing and planned IT platforms than ever before. This trend will accelerate over the next three years. CIOs and IT staffs need to start thinking about how they can become business strategists first, technicians and enablers of IT second. CIOs must create and launch new business models faster to keep their companies competitive. APIs are the fuel helping to make this happen.

The Urgency To Create New Business Models Is Driving API Proliferation

APIs (Application Programmer Interfaces) are the components that enable diverse platforms, apps, and systems to connect and share data with each other.  Think of APIs as a set of software modules, tools, and protocols that enable two or more platforms, systems and most commonly, applications to communicate with each other and initiate tasks or processes. APIs are essential for defining and customizing Graphical User Interfaces (GUIs) too. Cloud platform providers all have extensive APIs defined and work in close collaboration with development partners to fine-tune app performance using them. Amazon Web Services, Facebook, Google, Marketo, Salesforce, SAP Hybris, Twitter and thousands of other companies have APIs available. As of today, the Programmable Web lists 16,590 APIs in its database.

Removing The Hype By Benchmarking API Maturity .... "

Machine Learning as Service, Stealing the Model?

Interesting thoughts on constructing models. Algorithms for hire?  From Cornell:

Stealing Machine Learning Models via Prediction APIs

Florian Tramèr, Fan Zhang, Ari Juels, Michael K. Reiter, Thomas Ristenpart
(Submitted on 9 Sep 2016 (v1), last revised 3 Oct 2016 (this version, v2))
Machine learning (ML) models may be deemed confidential due to their sensitive training data, commercial value, or use in security applications. Increasingly often, confidential ML models are being deployed with publicly accessible query interfaces. ML-as-a-service ("predictive analytics") systems are an example: Some allow users to train models on potentially sensitive data and charge others for access on a pay-per-query basis. 

The tension between model confidentiality and public access motivates our investigation of model extraction attacks. In such attacks, an adversary with black-box access, but no prior knowledge of an ML model's parameters or training data, aims to duplicate the functionality of (i.e., "steal") the model. Unlike in classical learning theory settings, ML-as-a-service offerings may accept partial feature vectors as inputs and include confidence values with predictions.

 Given these practices, we show simple, efficient attacks that extract target ML models with near-perfect fidelity for popular model classes including logistic regression, neural networks, and decision trees. We demonstrate these attacks against the online services of BigML and Amazon Machine Learning. We further show that the natural countermeasure of omitting confidence values from model outputs still admits potentially harmful model extraction attacks. Our results highlight the need for careful ML model deployment and new model extraction countermeasures. ..."

Making AI more Human

This direction makes sense, have not looked at the details yet.    In TechRepublic:  Towards more human Watson APIs .  That's good    "  .... IBM Watson, Big Blue's cognitive computing division, announced Tuesday that it was expanding the availability of Watson APIs with three new tools: Tone Analyzer, Emotion Analysis, and Visual Recognition. IBM Watson is also updating their Text to Speech (TTS) feature and rebranding it as Expressive TTS .... " 

Understanding API Ecosystems

In the CACM:  Revealing the API Ecosystem and Enterprise Strategy via Visual Analytics.   Valuable insights can be gained by applying visual analytic techniques to understand complex, emerging ecosystem dynamics and evolving enterprise strategies. ... " 

API's for Machine Learning

An overview of a number of APIs for machine learning in Kdnuggets.  Also a look at the challenges involved for using this approach.  High level overviews of API's including those from IBM, Amazon, Google and Microsoft. I note that the ones in the IBM list are those included in the Watson portfolio. Worth consideration for using machine learning and decision implementation  inside your application.

Creating Cognitive Applications with the Watson API

Creating Cognitive Applications Powered by IBM Watson: Getting started with the API
by Kevin Haverlock, Sridhar Sudarsan
" ... Cognitive computing systems are built to learn from how they are used and adapt to their processing. They provide insight into content; offer context specific responses and guidance, and learn through iterative feedback cycles. While there are numerous ways to understand and extract meaningful information from structured data, the process of understanding and making sense of unstructured data is in its infancy. It requires systems to cull through and analyze complex and heterogeneous content sources to identify vague and abstruse information, patterns and relationships.

IBM Watson is at the forefront of cognitive computing. It furthers advances in natural language processing and analytics and provides the unique ability to comprehend the subtle nuances of human language, sift through vast amounts of content, and provide evidence-based responses to users' questions. Watson can be trained to provide responses with an increasing relevance over time. These capabilities, when access to them are embedded within an application, start them on a journey to building cognitive applications. .... "

Explore Data with Excel

A good introductory review.  There are still lots of groups that use Excel exclusively for manipulating data.  Now an add-in that addresses the further exploration of data.    Especially the conversion of data for explorative use.  " ...  Microsoft Data Explorer Preview for Excel is designed to help by easing data import from a variety of sources. You can then massage the data in different ways, from reshaping to filtering and grouping. It also simplifies merging tables on a common column ... " 

Body Imaging: Poikos

Stan Dyck sends along a link to Poikos,  " ...  Poikos means 3D body measurement, on any device, in seconds. Thanks for dropping by. After a successful closed beta period last year, we’ve opened our platform to developers world-wide. We’ll build the next phase of the Poikos platform together! Everything you need to know about our API has been released here. Did you see our call for collaborators? Please contact us directly for the SDK (web/smartphone/tablet). Why not give the latest version of our Flixfit iOS app a whirl as well? ... " 

Cloud As Innovation Enabler in Consumer Goods

A set of observations about what is going on in Consumer Products from Smarter Planet.  Some good examples of innovations to follow from the use of the Cloud.   " ... In fact, recent research by IBM indicates that while 16 percent of the global sample of business leaders surveyed were already using cloud capabilities for sweeping innovation (such as entering new lines of business or reshaping an existing industry) within the next three years, 35 percent intend to use it to transform their business models.

To explore this further, The Consumer Goods Forum has collaborated with IBM on a new study that looks at how cloud will drive innovation in the consumer goods space. In conjunction with IBM we presented a preview of the findings at our recent CGF Paris Conference which focused on marketing, IT and supply chain issues, and have now published the full study.  .... " 

Efficiency vs Value in Innovation

In Innovation Excellence:   A look at the innovation equation.  Measurement is always a useful exercise, even if you don't use the measure directly. " ... What organizations need to understand is that, at the macro level, productivity – which of course is central to profitable economic growth – has always been determined by two elements. On one side, it is determined by the efficiency with which companies use their inputs – how much labor and capital it takes to produce their goods and services. On the other side, productivity is determined by the value that customers place on the outputs.... " 

Changing Perceptions of Big Data

From Inside Analysis.  Here I disagree in one sense and agree in another. .  It is true that it has never entirely been about the size of the data.  What we have seen is that we are moving away from the hype that has been seen in the last several years.  As I have said here many times, it is about what you can do with new data, large or small.  It reminds me of the view of artificial intelligence in the late 80s.  Where a rapid change in perception caused a whole market, with valuable insights,  to collapse.  Good piece other than that assumption, and the perhaps unintended consequences.