Home Blog Page 303

Adobe launches subscriptions for Firefly AI

Avery Winslow
-
0
Adobe launches subscriptions for Firefly AI

Adobe Unveils Standalone Firefly AI Models Subscription Service

Adobe is launching a new standalone subscription service that gives users access to its AI image, vector, and video generating models. This marks the company’s boldest attempt yet to turn its Firefly AI models into a real product.

New Firefly AI Video Model

The company is also launching a redesigned webpage, firefly.adobe.com, where people can use Adobe’s AI models. This includes the new Firefly AI video model, which is rolling out in public beta on the Firefly website and in the Premiere Pro Beta app.

Firefly AI Models: What You Need to Know

The Firefly Standard plan costs $9.99 per month and provides unlimited access to Adobe’s AI image and vector generating features, as well as Adobe’s new AI video model. The Standard plan gives users 2,000 credits, which is enough to make 20 five-second AI videos.

Plans and Pricing

  • Firefly Standard: $9.99 per month, provides unlimited access to AI image and vector generating features, and new AI video model. 2,000 credits included (enough to make 20 five-second AI videos).
  • Firefly Pro: $29.99 per month, offers enough credits to generate 70 five-second AI videos per month.

Adobe’s Pitch to Creative Professionals

Part of Adobe’s pitch to creative professionals is that Firefly was trained on a dataset of licensed videos, without any brand logos or NSFW content. This means that creatives should be able to use the Firefly AI models without worrying about legal troubles.

Comparing Firefly to Other AI Video Models

The Firefly video model lets you turn text or images into a five-second, AI-generated video. There are controls on a side panel for changing the camera angles, camera movement, aspect ratio, and other features that creative professionals might want to customize.

The Competition

The new Firefly offerings will compete directly with OpenAI’s Sora, Runway’s Gen-3 Alpha, and other AI video models that already have dedicated webpages and subscription plans. Google DeepMind’s AI video model, Veo, is also a contender in the space, but it’s still in private beta.

Adobe’s Vision for Firefly

"We think the key differentiator for us is that we’re the only IP-friendly, commercially-safe video model," said Alexandru Costin, VP of Generative AI. "We want to differentiate with deep understanding of customer problems."

Adobe’s AI Tools for Creative Professionals

Adobe has also tried to ship AI tools that solve problems for creative professionals instead of just generating random AI videos. For example, one of Firefly’s AI video features, Generative Extend, lets users extend any clip’s video and background noise by a few seconds. This is one of the more practical AI video tools on the market.

Conclusion

Adobe is taking a bold step by launching a standalone subscription service for its Firefly AI models. The company is betting that creative professionals will be willing to pay for access to its AI image, vector, and video generating models. With its new Firefly AI video model, Adobe is trying to differentiate itself from other AI video models by offering a commercially-safe and IP-friendly solution.

FAQs

Q: What is Firefly?
A: Firefly is a new AI model from Adobe that can generate images, vectors, and videos.

Q: What is the cost of the Firefly subscription?
A: The Firefly Standard plan costs $9.99 per month, while the Firefly Pro plan costs $29.99 per month.

Q: What is the difference between Firefly Standard and Firefly Pro?
A: Firefly Standard provides unlimited access to AI image and vector generating features, as well as the new AI video model, with 2,000 credits. Firefly Pro offers enough credits to generate 70 five-second AI videos per month.

Q: Is Firefly IP-friendly and commercially-safe?
A: Yes, Firefly was trained on a dataset of licensed videos, without any brand logos or NSFW content, making it a safe and IP-friendly solution for creative professionals.

Hoarders, Collectors and Dumpster Divers

Phil
-
0
Hoarders, Collectors and Dumpster Divers

Hoarders, Collectors, and Dumpster Divers: Part 1

Introduction

This series showcases a selection of images from the ongoing "Hoarders, Collectors, and Dumpster Divers" series, highlighting the unique and fascinating world of hoarding and collecting.

Meet the Characters

Charlie

[Image: "Here’s Charlie"]

Johnny

[Image: "Here’s Johnny"]

Sammy

[Image: "Here’s Sammy"]

Emma

[Image: "Here’s Emma"]

Floss and Molly

[Image: "Here’s Floss and Molly"]

Salvia

[Image: "Here’s Salvia"]

Ethel and Pearl

[Image: "Here’s Ethel and Pearl"]

Tim and Bruce

[Image: "Here’s Tim and Bruce"]

Toby

[Image: "Here’s Toby"]

Conclusion

This selection of images provides a glimpse into the world of hoarding and collecting, showcasing the diverse and often extraordinary items that people collect and hoard. The series will continue to explore this fascinating topic, delving deeper into the motivations and habits of those who hoard and collect.

FAQs

Q: What is the purpose of the "Hoarders, Collectors, and Dumpster Divers" series?
A: The series aims to showcase the unique and fascinating world of hoarding and collecting, highlighting the motivations and habits of those who engage in these activities.

Q: What types of items are featured in the series?
A: The series features a wide range of items, including everyday objects, antiques, and unusual collectibles.

Q: How often will new images be added to the series?
A: New images will be added regularly, with the series continuing to explore different aspects of hoarding and collecting.

Q: Is this series intended to be humorous or serious?
A: The series is intended to be a thought-provoking and entertaining exploration of the world of hoarding and collecting, striking a balance between humor and serious reflection.

The billion-dollar AI company no one is talking about

Emerson Blake
-
0
The billion-dollar AI company no one is talking about

Understanding the AI Landscape

Before I jump into the who, I need to bring you up to speed.

As you know, there is so much hype and speculation about AI that DeepSeek made an announcement and the market crashed by trillions of dollars. Yes, trillion with a T. When I say winning, I mean winning. Not "potentially" winning. They are already winning.

The Big Winner

Yup. That Reddit. I told you this was a Martin Scorsese plot twist.

If you’re unfamiliar with Reddit, it is a social media platform and discussion site where users share content, discuss topics, and vote on posts. It’s often called "The Front Page of the Internet."

Why Reddit is Winning

Here is why they are winning… AI models need human-generated content for training. Reddit just so happens to have one of the largest human conversation datasets in the world. What makes this so valuable is the quality of the conversations. The internet is full of junk, but Reddit’s discussions are real, unfiltered, and human. Unlike algorithm-driven platforms, this is raw, natural interaction, which is precisely what AI needs to understand human behavior.

The Opportunity!

Reddit has become crucial for AI development, and no one else comes close to having this level of rich, human-generated data. It’s seen as the last town square for free human conversation.

Here is your opportunity.

Conclusion

Reddit has proven to be a winner in the AI landscape, and its newfound profitability means it will likely invest more in the platform to improve user experience, add new features, and scale its reach. As a marketer, this is an opportunity to connect with the right audience and provide value.

Frequently Asked Questions

Q: What is Reddit?
A: Reddit is a social media platform and discussion site where users share content, discuss topics, and vote on posts.

Q: Why is Reddit winning in AI?
A: AI models need human-generated content for training, and Reddit has one of the largest human conversation datasets in the world.

Q: What is the opportunity for marketers?
A: Marketers can connect with the right audience and provide value on Reddit, a platform that is growing rapidly and becoming more profitable.

Q: How can I get started with Reddit marketing?
A: Start by understanding the community, providing value, and focusing on authentic human interaction.

Wacom’s New Intuos Pro Drawing Tablet

Avery Winslow
-
0
Wacom’s New Intuos Pro Drawing Tablet

Wacom Intuos Pro: A Redesigned Digital Drawing Tablet for Creative Professionals

Overview

Wacom has overhauled its Intuos Pro line of digital drawing tablets, introducing a new design that is more compact, convenient, and customizable for creative professionals. The redesign features a new layout that relocates user controls from the side to the top of the tablet and replaces the previous touch ring with mechanical dials.

New Design and Features

The Intuos Pro is designed for industry creatives like 3D artists, illustrators, and photo editors who prefer to work on their own displays. Unlike Wacom’s Cintiq or One range, the Intuos Pro does not have a built-in screen, making it easier to carry and connect to computing devices. The tablet is available in three sizes: Small (€269.99), Medium (€399.99), and Large (€549.99).

Smaller and More Compact

Each model is actually smaller than its predecessor, yet features a larger active area that makes better use of the space. For example, the 2025 Medium measures 11.5 x 8.1 inches, with a drawing area of 10.4 x 5.8 inches, compared to 8.7 x 5.8 inches for the previous model.

Redesigned User Controls

The most noticeable change is the new layout, where user controls like customizable buttons and touch-sensitive wheels have been shifted to the top of the tablet. This reduces the space on either side of the active pen area and makes it easier for left-handed or ambidextrous users to configure for their workflows.

Mechanical Dials and Customizable Express Keys

The touch wheel has been replaced by tactile mechanical dials, providing more precision when scrolling through tools and actions. The customizable express keys have also been updated to a round space-saving design, with five buttons on the Small and ten on the two larger models.

Wireless Connectivity and Compatibility

The Intuos Pro works with devices running Windows 10 or macOS 13 or later, supporting both wired USB-C to USB-A connection and wireless via Bluetooth 5.3. The latter provides up to 16 hours of battery life when fully charged, and users can switch between USB and Bluetooth via a switch on the side of the tablet, allowing it to pair with two devices simultaneously.

Wacom Pro Pen 3 and Compatibility

The Intuos Pro comes with the same customizable Wacom Pro Pen 3 stylus, which features adjustable thickness and buttons. The Pro Pen 3 can switch between devices, making it useful for working across multiple tablets. The new Intuos Pro is also compatible with previous-gen Wacom Pro Pens and EMR styluses from Pilot, Dr. Grip, LAMY, and Staedtler.

Initial Impressions

I was able to handle a pre-production sample for a short time, and while a full review will follow, the slimmer design felt well-constructed and wasn’t flimsy. The new layout is a welcome improvement, and I didn’t take long to adjust to the tactile mechanical dial.

Conclusion

The Wacom Intuos Pro is a significant upgrade for creative professionals, offering a more compact, convenient, and customizable digital drawing experience. With its new design and features, it’s an excellent option for those looking to upgrade their digital art workflow.

FAQs

Q: What devices is the Wacom Intuos Pro compatible with?
A: The Intuos Pro works with devices running Windows 10 or macOS 13 or later.

Q: What are the different sizes of the Wacom Intuos Pro available?
A: The Intuos Pro is available in three sizes: Small (€269.99), Medium (€399.99), and Large (€549.99).

Q: What is the battery life of the Intuos Pro?
A: The Intuos Pro has a battery life of up to 16 hours when fully charged, using wireless connectivity via Bluetooth 5.3.

Q: Is the Wacom Pro Pen 3 compatible with other Wacom devices?
A: Yes, the Wacom Pro Pen 3 can switch between devices, making it useful for working across multiple tablets.

Disney’s Animation Future in Question

Phil
-
0
Disney’s Animation Future in Question

Disney Revamps DEI Initiatives: What’s Changing and How It Affects Animation

Changes to Disney’s DEI Approach

Disney is revamping its Diversity, Equity, and Inclusion (DEI) initiatives, making significant changes to its approach to diversity, equity, and inclusion. This shift follows a trend among companies, including federal agencies, in response to US President Donald Trump’s executive order targeting DEI programs.

Internal Changes

Internally, Disney is replacing Diversity & Inclusion performance ratings with a new Talent Strategy category. The most notable change is the cancellation of the Reimagine Tomorrow initiative, designed to promote underrepresented voices.

Impact on Character Design and Storytelling

The removal of content advisory warnings from classic films on Disney+ will be a visible change. However, the deeper impact will be on character design, talent promotion, and the types of stories Disney can now tell. This shift may affect the kinds of stories and characters Disney creates, potentially limiting the representation of underrepresented groups.

Recent Efforts to Promote Diversity and Inclusion

In recent years, Disney has made significant strides in promoting diversity and inclusion. Notable examples include:

  • The Princess and the Frog
  • Halle Bailey’s casting as Ariel in The Little Mermaid
  • Pixar’s Elemental, featuring a non-binary character
  • Turning Red, focusing on a Chinese-Canadian girl’s journey
  • Moana, Raya and the Last Dragon, and Strange World, which feature diverse characters and stories

The Future of Animation

Will Disney’s shift away from DEI initiatives spell the end of original and inventive storytelling and character design? The answer lies in the company’s ability to adapt and evolve while maintaining its commitment to diversity and inclusion.

Conclusion

Disney’s revamp of its DEI initiatives may signal a new era for the company, one that prioritizes talent strategy over diversity and inclusion. While this change may have significant implications for character design and storytelling, it is essential to remember the progress made in recent years. The future of animation remains uncertain, but the impact on Disney’s commitment to diversity and inclusion will be closely watched.

FAQs

Q: What does the removal of content advisory warnings from classic films on Disney+ mean?
A: It is a visible change that may indicate a shift away from promoting diversity and inclusion.

Q: What is the significance of the cancellation of the Reimagine Tomorrow initiative?
A: It was designed to promote underrepresented voices, and its cancellation may limit opportunities for underrepresented groups to be represented in Disney’s content.

Q: How will this change affect Disney’s character design and storytelling?
A: The shift may lead to a decrease in representation of underrepresented groups, potentially limiting the stories and characters Disney creates.

Q: What does the future of animation look like for Disney?
A: The future is uncertain, and it is essential to monitor the company’s commitment to diversity and inclusion to understand the impact on character design and storytelling.

Hoarders, Collectors and Dumpster Divers

Phil
-
0
Hoarders, Collectors and Dumpster Divers

Hoarders, Collectors, and Dumpster Divers: Part 2

Featured Individuals

Bernie


"Here’s Bernie"

Trevor


"Here’s Trevor"

Yukito and Sharleeta


"Here’s Yukito and Sharleeta"

Bruski


"Here’s Bruski"

Gede and Seti


"Here’s Gede and Seti"

The Rassee Sangraahak Family


"Here’s the Rassee Sangraahak Family"

Conrad


"Here’s Conrad"

Evelyn and Ardan


"Here’s Evelyn and Ardan"

The Duckworths


"Here’s the Duckworths"

Shannon Mitchell


"Here’s Shannon Mitchell"

Dr. Bonŝancon


"Here’s Dr. Bonŝancon"

Shannon Mitchell Again


"Here’s Shannon Mitchell Again"

Mia and Antonio


"Here’s Mia and Antonio"

Teddy


"Here’s Teddy"

Julia and Kyle


"Here’s Julia and Kyle"

Conclusion

This selection of images showcases the diverse range of individuals featured in the "Hoarders, Collectors, and Dumpster Divers" series. Each individual has their own unique story and collection, and this series aims to highlight the fascinating world of collecting and hoarding.

FAQs

Q: What is the purpose of the "Hoarders, Collectors, and Dumpster Divers" series?
A: The series aims to showcase the diverse range of individuals who collect and hoard, and to highlight the fascinating world of collecting and hoarding.

Q: How do you choose the individuals featured in the series?
A: The individuals are chosen based on their unique collecting habits and stories, and are selected to showcase the diversity and richness of the world of collecting and hoarding.

Q: Can I submit my own collection or hoard for consideration for the series?
A: Yes, please contact us with your story and images, and we will review them for consideration.

Falcon 3 models now available in Amazon SageMaker JumpStart

Jordan Waverly
-
0
Falcon 3 models now available in Amazon SageMaker JumpStart

Overview of the Falcon 3 family of models

The Falcon 3 family, developed by Technology Innovation Institute (TII) in Abu Dhabi, represents a significant advancement in open source language models. This collection includes five base models ranging from 1 billion to 10 billion parameters, with a focus on enhancing science, math, and coding capabilities. The family consists of Falcon3-1B-Base, Falcon3-3B-Base, Falcon3-Mamba-7B-Base, Falcon3-7B-Base, and Falcon3-10B-Base along with their instruct variants.

Get started with SageMaker JumpStart

SageMaker JumpStart is a machine learning (ML) hub that can help accelerate your ML journey. With SageMaker JumpStart, you can evaluate, compare, and select pre-trained foundation models (FMs), including Falcon 3 models. These models are fully customizable for your use case with your data.

Deploy Falcon 3 using the SageMaker JumpStart UI

To access SageMaker JumpStart, use one of the following methods:

  • In Amazon SageMaker Unified Studio, on the Build menu, choose JumpStart models under Model development.
  • Alternatively, in Amazon SageMaker Studio, choose JumpStart in the navigation pane.

Search for Falcon3-10B-Base in the model browser. Choose the model and choose Deploy. For Instance type, either use the default instance or choose a different instance. Choose Deploy. After some time, the endpoint status will show as InService and you will be able to run inference against it.

Deploy Falcon 3 programmatically using the SageMaker Python SDK

For teams looking to automate deployment or integrate with existing MLOps pipelines, you can use the SageMaker Python SDK:


from sagemaker.serve.builder.model_builder import ModelBuilder
from sagemaker.serve.builder.schema_builder import SchemaBuilder
from sagemaker.jumpstart.model import ModelAccessConfig
from sagemaker.session import Session
import logging

sagemaker_session = Session()

artifacts_bucket_name = sagemaker_session.default_bucket()
execution_role_arn = sagemaker_session.get_caller_identity_arn()

js_model_id = "huggingface-llm-falcon-3-10B-base"

gpu_instance_type = "ml.g5.12xlarge"  

response = "Hello, I'm a language model, and I'm here to help you with your English."

sample_input = {
    "inputs": "Hello, I'm a language model,",
    "parameters": {"max_new_tokens": 128, "top_p": 0.9, "temperature": 0.6},
}

sample_output = [{"generated_text": response}]

schema_builder = SchemaBuilder(sample_input, sample_output)

model_builder = ModelBuilder(
    model=js_model_id,
    schema_builder=schema_builder,
    sagemaker_session=sagemaker_session,
    role_arn=execution_role_arn,
    log_level=logging.ERROR
)

model= model_builder.build()

predictor = model.deploy(model_access_configs={js_model_id:ModelAccessConfig(accept_eula=True)}, accept_eula=True) 

Run inference on the predictor:

predictor.predict(sample_input) 

If you want to set up the ability to scale down to zero after deployment, refer to Unlock cost savings with the new scale down to zero feature in SageMaker Inference.

**Clean up**

To clean up the model and endpoint, use the following code:

predictor.delete_model()
predictor.delete_endpoint()

**Conclusion**

In this post, we explored how SageMaker JumpStart empowers data scientists and ML engineers to discover, access, and run a wide range of pre-trained FMs for inference, including the Falcon 3 family of models. Visit SageMaker JumpStart in SageMaker Studio now to get started. For more information, refer to SageMaker JumpStart pretrained models, Amazon SageMaker JumpStart Foundation Models, and Getting started with Amazon SageMaker JumpStart.

**About the authors**

[Niithiyn Vijeaswaran](https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2024/05/30/NIITHIYN_ARP-2-1.jpg) is a Generative AI Specialist Solutions Architect with the Third-Party Model Science team at AWS. His area of focus is generative AI and AWS AI Accelerators. He holds a Bachelor’s degree in Computer Science and Bioinformatics.

[Marc Karp](https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2024/11/27/ml-17555-karpmar.jpg) is an ML Architect with the Amazon SageMaker Service team. He focuses on helping customers design, deploy, and manage ML workloads at scale. In his spare time, he enjoys traveling and exploring new places.

[Raghu Ramesha](https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2024/11/27/ml-17555-raghu.png) is a Senior ML Solutions Architect with the Amazon SageMaker Service team. He focuses on helping customers build, deploy, and migrate ML production workloads to SageMaker at scale. He specializes in machine learning, AI, and computer vision domains, and holds a master’s degree in Computer Science from UT Dallas. In his free time, he enjoys traveling and photography.

[Banu Nagasundaram](https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2020/12/01/Banu-Nagasundaram.jpg) leads product, engineering, and strategic partnerships for SageMaker JumpStart, SageMaker’s machine learning and GenAI hub. She is passionate about building solutions that help customers accelerate their AI journey and unlock business value.

AI System Makes Lip Reading More Personal and Accurate

Parker Lennox
-
0
AI System Makes Lip Reading More Personal and Accurate

Overview

  • Novel system for personalized lipreading using audio-visual self-distillation
  • Adapts to individual speakers through specialized pretraining
  • Combines visual and audio data to improve accuracy
  • Introduces speaker adaptation techniques for better performance
  • Demonstrates significant improvements over traditional lipreading methods

Plain English Explanation

Think of lipreading like learning to understand someone’s unique way of speaking by watching their mouth movements. Just as we get better at understanding a friend’s speech patterns over time, this system learns to recognize specific individuals’ lip movements more accurately.

How it Works

The system uses a novel technique called audio-visual self-distillation, which combines visual and audio data to improve accuracy. This is achieved through a process called speaker adaptation, where the system is trained on a specific speaker’s voice and lip movements to better understand their unique characteristics.

Advantages

The system’s ability to adapt to individual speakers and combine visual and audio data makes it more accurate than traditional lipreading methods. This technology has the potential to revolutionize the way we communicate, particularly for those with hearing impairments or in noisy environments.

Conclusion

The AI system for target speaker lipreading by audio-visual self-distillation is a significant breakthrough in the field of lipreading. Its ability to adapt to individual speakers and combine visual and audio data makes it a more accurate and effective method for understanding lip movements. This technology has the potential to improve communication for individuals with hearing impairments and those in noisy environments.

FAQs

Q: How does the system work?

A: The system uses a novel technique called audio-visual self-distillation, which combines visual and audio data to improve accuracy.

Q: What is speaker adaptation?

A: Speaker adaptation is a process where the system is trained on a specific speaker’s voice and lip movements to better understand their unique characteristics.

Q: How is the system more accurate than traditional lipreading methods?

A: The system’s ability to adapt to individual speakers and combine visual and audio data makes it more accurate than traditional lipreading methods.

Q: What are the potential applications of this technology?

A: This technology has the potential to revolutionize the way we communicate, particularly for those with hearing impairments or in noisy environments.

Building A.I. with Less Money

Avery Winslow
-
0
Building A.I. with Less Money

How DeepSeek Built One of the World’s Most Powerful AI Systems with Fewer Computer Chips

How are A.I. technologies built?

The leading A.I. technologies are based on what scientists call neural networks, mathematical systems that learn their skills by analyzing enormous amounts of data. The most powerful systems spend months analyzing just about all the English text on the internet as well as many images, sounds, and other multimedia. That requires enormous amounts of computing power.

About 15 years ago, A.I. researchers realized that specialized computer chips called graphics processing units, or GPUs, were an effective way of doing this kind of data analysis. Companies like the Silicon Valley chipmaker Nvidia originally designed these chips to render graphics for computer video games. But GPUs also had a knack for running the math that powered neural networks.

How was DeepSeek able to reduce costs?

It did many things. Most notably, it embraced a method called "mixture of experts." Companies usually created a single neural network that learned all the patterns in all the data on the internet. This was expensive, because it required enormous amounts of data to travel between GPU chips.

With the mixture of experts method, researchers tried to solve this problem by splitting the system into many neural networks: one for poetry, one for computer programming, one for biology, one for physics, and so on. There might be 100 of these smaller "expert" systems. Each expert could concentrate on its particular field.

How is this efficient?

Many companies have struggled with this method, but DeepSeek was able to do it well. Its trick was to pair those smaller "expert" systems with a "generalist" system. The experts still needed to trade some information with one another, and the generalist – which had a decent but not detailed understanding of each subject – could help coordinate interactions between the experts.

And that is more efficient?

Much more. But that is not the only thing DeepSeek did. It also mastered a simple trick involving decimals that anyone who remembers his or her elementary school math class can understand.

There is math involved in this?

Remember your math teacher explaining the concept of pi. Pi, also denoted as π, is a number that never ends: 3.14159265358979 …. You can use π to do useful calculations, like determining the circumference of a circle. When you do those calculations, you shorten π to just a few decimals: 3.14. If you use this simpler number, you get a pretty good estimation of a circle’s circumference.

DeepSeek did something similar – but on a much larger scale – in training its A.I. technology. The math that allows a neural network to identify patterns in text is really just multiplication – lots and lots and lots of multiplication. We’re talking months of multiplication across thousands of computer chips.

That’s it?

Well, they added another trick. After squeezing each number into 8 bits of memory, DeepSeek took a different route when multiplying those numbers together. When determining the answer to each multiplication problem – making a key calculation that would help decide how the neural network would operate – it stretched the answer across 32 bits of memory. In other words, it kept many more decimals. It made the answer more precise.

So any high school student could have done this?

Well, no. The DeepSeek engineers showed in their paper that they were also very good at writing the very complicated computer code that tells GPUs what to do. They knew how to squeeze even more efficiency out of these chips.

Then why didn’t they do this already?

Some A.I. labs may be using at least some of the same tricks already. Companies like OpenAI do not always reveal what they are doing behind closed doors. But others were clearly surprised by DeepSeek’s work. Doing what the start-up did is not easy. The experimentation needed to find a breakthrough like this involves millions of dollars – if not billions – in electrical power.

Conclusion

DeepSeek’s innovative approach to building A.I. systems has the potential to significantly reduce the cost of building A.I. technologies. By embracing the mixture of experts method and using simple mathematical tricks, the start-up has shown that it is possible to build powerful A.I. systems with fewer computer chips. This breakthrough could have far-reaching implications for the development of A.I. technologies.

FAQs

Q: Can any high school student do this?
A: No, the DeepSeek engineers have a deep understanding of computer code and were able to write the complicated code that tells GPUs what to do.

Q: Why didn’t other A.I. labs do this already?
A: The experimentation needed to find a breakthrough like this involves millions of dollars – if not billions – in electrical power. It requires enormous amounts of risk, which can be a barrier to innovation.

Q: Will this breakthrough lead to cheaper A.I. systems?
A: Yes, DeepSeek’s innovative approach has the potential to significantly reduce the cost of building A.I. systems.

Why Defying Gravity was Wicked’s Hardest Scene to Make

Phil
-
0
Why Defying Gravity was Wicked’s Hardest Scene to Make

Making the VFX of Wicked: The Defying Gravity Sequence

The Pivotal Moment

Defying Gravity is an iconic song, and its scene in the movie adaptation of Wicked is crucial to the entire film. Getting this right played on the minds of the film’s VFX team at ILM, as well as the filmmakers, including director John M. Chu. In an exclusive interview, VFX supervisor for ILM, Pablo Helman, reveals what it took to make sure Defying Gravity, well… defied expectations.

Building Emerald City

The work began with building Emerald City. "We started with references from production design and usually the way that it works, at least for me, right after the shoot we download everything that production has done as a reference for everything that we do. We start looking at that stuff and we take it from there. So that gave us a language for the movie that we were making," says Pablo.

Cynthia Erivo’s Commitment

Some of what happened next was dictated by lead actress Cynthia Erivo and her commitment to the role. "After that, Cynthia Erivo wanted to do her own (flying) stunts and sing at the same time. That’s really difficult," says Pablo. "In a sequence that has to be made from the vision of the director, we worked on previz for months but because this is also a musical everything is tied to a specific sandbox and the fact that they are singing live opens the sandbox a little bit."

The Challenging Process

The Defying Gravity sequence was one of the most difficult sequences the team worked on. "We had to make someone fly, and we’ve seen it before, so how do we do it differently and how do we allow for that tactile process, that organic process, for Jon and the actors to feel part of what we’re making?" says Pablo.

The Aesthetics and Practical Connection

Pablo defines the aesthetic and practical connection between the Defying Gravity sequence and the larger effort. "There are always editorial needs, meaning that the team needs to move an actor from one place to another, but it may not happen on the shoot because it cannot physically be done. So, all of a sudden, the actor is moving at 700 miles an hour in the CG space. You have to have materials and lighting. At the end of the movie, Elphaba becomes bigger than life: there is an almost religious or spiritual sense to that character."

Creative Questions

Creative questions ensued for Pablo and his collaborators. "How are we going to mitigate her cape growing in the middle of a sequence where she’s flying around and where, physically, through the sequence, the cape is five or six feet long? When we got to that shot, that was when we had to suspend disbelief, right?"

Conclusion

The Defying Gravity sequence is a memorable moment in Wicked, thanks to ILM’s work and the dedication of the entire team. Inspired by ILM’s work? Then read our guide to the best 3D software and the best animation software for an insight into the tools VFX artists use.

Frequently Asked Questions

Q: What was the most challenging part of the Defying Gravity sequence?
A: Making someone fly and allowing for the tactile process for the actors to feel part of what we’re making.

Q: How did the VFX team approach the sequence?
A: We started with references from production design and worked on previz for months.

Q: What was the goal of the Defying Gravity sequence?
A: To make someone fly and to allow for the tactile process for the actors to feel part of what we’re making.

Q: How did the VFX team overcome the challenges of the sequence?
A: By using a combination of previz, materials, and lighting to create a believable and memorable experience.

1...302303304...741Page 303 of 741

Welcome to MachineDaily.AI, your go-to destination for breaking news, in-depth analysis, and the latest developments in the world of artificial intelligence (AI). Our mission is to provide you with timely and insightful content that explores AI’s vast potential across industries and its growing impact on society. From cutting-edge innovations to ethical debates, MachineDaily.AI is committed to covering AI in all its forms, making complex topics accessible to experts, enthusiasts, and curious minds alike.

Latest

Categories

Useful Links

Our Newsletter

Subscribe Us To Receive Our Latest News Directly In Your Inbox!

© Copyright 2024. All Right Reserved By Machine Daily.