Author: Oriol Zertuche

Oriol Zertuche is the CEO of CODESM and Cody AI. As an engineering student from the University of Texas-Pan American, Oriol leveraged his expertise in technology and web development to establish renowned marketing firm CODESM. He later developed Cody AI, a smart AI assistant trained to support businesses and their team members. Oriol believes in delivering practical business solutions through innovative technology.

Can the SAP Generative AI “Joule” Be Your Business Copilot?

Posted on October 10, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence

Joule is designed to generate responses based on real-world situations. The German multinational software giant is putting in the effort to make sure Joule is not just productive but also ethical and responsible. They're gearing up for a future where generative AI plays a central role in personal and professional settings.

Recognizing the increasing prevalence of generative AI in daily life, SAP generative AI assistant, a business copilot named Joule, is here! It’s intriguing to see how generative AI is gaining ground in different parts of the world. About half of Australians surveyed, about 49%, utilize generative AI. In the US, it’s 45%; in the UK, it’s 29%.

What is SAP Generative AI Joule?

Joule is designed to generate responses based on real-world situations. The German multinational software giant is putting in the effort to make sure Joule is not just productive but also ethical and responsible. They’re gearing up for a future where generative AI plays a central role in personal and professional settings.

Joule is going to be part of all SAP applications. Joule will be right there whether you’re dealing with HR, Finance, Supply Chain, or Customer Experience.

What’s it all about?

Well, imagine being able to ask a question or put a problem in plain language and get intelligent, context-preserving replies.

That’s precisely what Joule brings to the table. It taps into extensive business data from SAP’s comprehensive portfolio and outside sources to ensure you get the most insightful and relevant answers.

Consider that you’re facing a challenge: determining ways to improve your logistics processes. To present viable solutions to the manufacturer for assessment, Joule can spot regions where your sales might be underperforming.

On top of that, it can connect to other data sets that hint at a supply chain issue and instantly connect to the supply chain system. But it doesn’t stop there. Joule is a versatile assistant, and it’s there for you across all SAP applications, continually supplying new situations.

What Makes Joule a Top-Class SAP Generative AI Assistant?

Being one of the world’s leading enterprise resource planning software vendors, SAP takes data protection and fairness seriously. One of the standout features is its commitment to keeping biases out of the Large Language Models (LLMs) Joule deploys.

Increased Efficiency

Source

Enhance your productivity with an AI assistant that understands your specific role and collaborates seamlessly within the SAP applications, streamlining your tasks.

Enhanced Intelligence

Access rapid responses and intelligent insights whenever you need them, enabling faster decision-making without workflow interruptions.

Improved Results

Source

Simply inquire, and receive customized content to kickstart your tasks. Generate job descriptions, obtain coding guidance, and more with ease.

Total Autonomy

Source

Retain complete control over your decision-making and data privacy while utilizing generative AI in a secure and controlled environment.

Joule won’t train LLMs using customer information. Your data stays safe, and there’s no risk of unintentional bias creeping into the AI’s responses.

SAP’s Generative AI Assistant’s Rollout Plan

The rollout of Joule is happening in stages across SAP’s suite of solutions. Here’s what you can expect:

Later this year, Joule will debut with SAP SuccessFactors solutions and become accessible through the SAP Start site.
Next year, it will expand its reach to SAP S/4HANA Cloud, public edition. So, if you’re using that, Joule will be there to assist.
Beyond that, Joule will continue its journey and become an integral part of SAP Customer Experience and SAP Ariba solutions.
It will also join the SAP Business Technology Platform, ensuring it’s available across a wide range of SAP applications.

So, Joule’s on the move, gradually making its way into different corners of the SAP ecosystem to enhance your experiences.

What to Expect from SAP Generative AI Joule?

Uncertainty exists around pricing. According to SAP’s previous projections, embedded AI for business capabilities might bring in a 30% premium. But the good news is that some of Joule’s features will be available to customers without extra cost. On the other hand, for certain advanced capabilities tailored to specific business needs, a premium might be involved. So, it depends on how you plan to use it.

Conclusion

As a generative AI assistant, Joule is poised to revolutionize business operations with its intelligent responses and problem-solving across SAP applications.

With SuccessConnect on October 2–4, Spend Connect Live on October 9–11, Customer Experience LIVE on October 25, the SAP TechEd conference on November 2–3, and many more, keep your calendars marked because SAP has a whole lineup of exciting updates coming your way!

Meet LLaVA: The New Competitor to GPT-4 Vision

Posted on October 7, 2023 by Oriol Zertuche - AI tools, Design

OpenAI’s GPT-4 image recognition technology recently took the tech world by storm. However, even as the dust was settling, a new contender has entered the fray: LLaVA, or the Large Language and Vision Assistant. Open-sourced and absolutely free to use, LLaVA is set to redefine the boundaries of image recognition technology.

What is LLaVA?

LLaVA is a cutting-edge tool created by experts from University of Wisconsin-Madison, Microsoft Research, and Columbia University. In simple terms, it’s a piece of technology designed to understand both visuals (like photos) and language (like text). Just imagine ChatGPT that can chat about a picture as well as a human can, and that’s LLaVA for you.

Why is LLaVA Special?

LLaVA isn’t just another image recognition tool. It blends a vision “encoder” (think of this as the eyes of the system) with something called Vicuna (its brain for understanding language). This combo makes LLaVA a superstar in chatting about images and understanding complex visual info, just like how GPT-4 Vision does.

Open-Source and Ready to Use

What’s even more exciting? If you’re a tech enthusiast or a developer, you can dive into LLaVA’s inner workings. The creators have kindly shared everything online. From its blueprint (or Paper) to the actual Code and Model, it’s all out there for those curious minds.

In Conclusion

While the image recognition technology landscape is fiercely competitive, LLaVA has undoubtedly carved a niche for itself in a short span. Its remarkable performance, combined with its open-source nature, make it a force to reckon with in the tech world.

The age of image recognition technology is evolving rapidly, and with LLaVA now in the mix, the future looks even more promising. The only question is: Are you ready to be a part of this visual revolution?

Discover More about LLaVA

How to Create Vector Illustrations with DALL-E 3 Prompts

Posted on October 7, 2023 by Oriol Zertuche - Design, Marketing

DALL-E 3 prompts can help generate highly detailed and realistic images. To do this, it assesses a broad image database and crafts relevant visuals with the help of the input text.

The world of graphic design has witnessed a dramatic transformation in recent years. There was a time when creating vector images was a time-consuming process, requiring extensive knowledge of complex software like Adobe Illustrator. It would often take hours, even for a professional, to perfect a single vector design.

But today, things have changed drastically. With the advent of AI technology, the process of creating vector images has been simplified and democratized. ChatGPT Plus, an advanced AI model developed by OpenAI, has played a pivotal role in this transformation.

This powerful tool is capable of understanding text descriptions and generating vector images based on them. It’s as simple as typing a detailed prompt and letting the AI do its magic. The more specific your description, the closer the output will be to your vision.

But how does one use DALL-E 3 prompts to create vector images from text?

Is it as straightforward as entering a description and waiting for the image to materialize?

In essence, yes. However, there are some nuances to consider if you want to optimize the results. And that’s exactly what we’re going to explore in this blog. Let’s dive in.

Step 1: Envision Your Image

The initial step in the process is to define the image you want to generate that aligns with your business needs. It’s important to remember that if you’re too broad with your description, the AI may create a more complex environment which could detract from the vector style look you’re aiming for. So, being specific about the items and elements you want present in the scene can help maintain the desired style.

For instance, instead of just stating “a business meeting”, you might say “a man and women professionals brainstorming with a white board behind them”. The more specific you are in your description, the better the AI can match your vision.

Step 2: Design Your DALL-E 3 Prompt

Once you’ve identified your concept, it’s time to formulate your prompt. Your prompt should include both the subject matter (what you want to see) and stylistic elements (how it should look). For this scenario, your style modifiers could be “flat simple vector illustrations style, vibrant colors, white background”.

Here’s a complete DALL-E 3 prompt example:

a man and women professionals brainstorming with a white board behind them, flat simple vector illustrations style, vibrant colors, white background

Copy To Clipboard

Step 3: Generate Your Image

With your DALL-E 3 prompt ready, you can now generate your image. The beauty of this AI tool is its ability to translate your text prompt into a unique visual representation. However, your work doesn’t stop here. After generating the image, you can further refine it by iterating on your prompt or using the AI’s edit features. You can tweak certain elements, add additional details, or even change the style until you get the image that perfectly matches your vision.

If you want to create a series of coherent images, consider adding a color theme to your prompt. For example, if you’re creating a series of images for a presentation, you might want to stick to a “blue theme” to ensure visual consistency. Simply add your color theme before the background description in your AI prompt. This way, as you generate and refine your images, they will maintain a consistent color scheme that aligns with your business branding or presentation theme.

Step 4: Convert Your DALL-E 3 Image to SVG

After you’ve created and refined your image to your satisfaction, you might want to convert it into a scalable vector graphics (SVG) format. SVG is a popular format for vector images, as it allows for high-quality visuals at any scale without pixelation.

To do this, you can use Adobe Express SVG. It’s a free tool that allows you to easily upload your image and convert it into an SVG file. Simply upload your image, wait for the conversion to complete, and then download your new SVG file. It’s a straightforward process that makes it easy to incorporate your AI-generated art into various types of content.

And with that final step, we’ve reached the end of our journey, but it’s really just the beginning of yours.

Read More: DALL-E 3 for Marketing

Conclusion

With the advent of AI technologies like Dall-E 3, creating vector images has been transformed from a time-consuming task to an exciting and creative journey. The process we’ve explored today – envisioning your image, designing your prompt, generating and refining your image, and converting your final product to SVG format – is just the start. With these tools and techniques, you now have the power to bring your unique visions to life. So, go forth and create, and remember, the only limit is your imagination.

Microsoft Copilot: The Latest AI in Business

Posted on October 2, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence

Microsoft Copilot has been meticulously architected to uphold the standards of security, compliance, and privacy. It is integrated into the Microsoft 365 ecosystem.

Imagine having a virtual assistant right there in your Microsoft 365 apps, like Word, Excel, PowerPoint, Teams, and more. As AI in business, Microsoft Copilot is here to make your work lives easier and more efficient. Let’s find out what it’s all about!

Microsoft Copilot’s Impact on Your Daily Workflows

Think about it: you’re in a meeting and turn to Microsoft Copilot for answers related to the agenda. What happens next is that Copilot doesn’t just give you a generic response; it brings together insights from past email exchanges, documents, and chat discussions. It’s like it remembers every detail, all rooted in your unique business context.

Microsoft Copilot in Action Across Apps

Microsoft Copilot is designed to be your collaborator, integrated into Word, Excel, PowerPoint, Outlook, Teams, or other Microsoft 365 apps you use daily. Whether you’re using Outlook to write emails or working on a presentation in PowerPoint, Copilot offers a shared design language for prompts, refinements, and commands.

But Copilot’s capabilities don’t end there. It can command apps, enabling actions like animating a slide, and it’s proficient at working across applications, effortlessly translating a Word document into a PowerPoint presentation.

Integration With Business Chat: A Game-Changer for Workplace Efficiency

Another key component of Copilot’s integration is through Business Chat, which operates across LLMs (Large Language Models), Microsoft 365 apps, and your own data. Copilot can perform various NLP (Natural Language Processing) tasks thanks to its deep learning algorithm. Moreover, the integration gives real-time access to your business content—think documents, emails, calendars, chats, meetings, and contacts.

This combination of your data with your immediate working context, whether it’s your meeting, emails you’ve exchanged, or chat convos from last week, leads to precise and contextual responses. Microsoft 365 Copilot streamlines your workflow and improves your skill set, making your work life smoother, more creative, and way more efficient.

A Foundation of Trust

Microsoft Copilot has been meticulously architected to uphold the standards of security, compliance, and privacy. It is integrated into the Microsoft 365 ecosystem. So, Copilot naturally follows your organization’s security and privacy rules, whether it’s two-factor authentication, compliance boundaries, or privacy safeguards.

The Power to Learn and Adapt

Copilot is designed to be a continuous learner. It adapts and learns new skills when it faces new domains and processes. For instance, with Viva Sales, Copilot can learn to connect with customer relationship management (CRM) systems. It can pull in customer data, such as interaction and order histories, and incorporate this information into your communications.

Copilot’s knack for continuous learning ensures that it won’t stop at ‘good’; it will aim for ‘exceptional’ as it evolves, becoming even more precise and capable over time.

Conclusion

The future of work is here, and it’s called Microsoft 365 Copilot. Leveraging LLMs and integrating them with your business data, Copilot transforms your everyday apps into something extraordinary, unlocking many amazing possibilities.

Copilot supercharges your productivity, always understands the context, keeps your data safe, and offers a consistent experience. Plus, it’s a quick learner, adapting to your business needs. With Copilot by your side, the future of work looks more intelligent and efficient than ever!

Meta AI’s Latest Launch: All You Need to Know

Posted on October 1, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence

Meta's latest hardware innovations at the Connect conference, featuring the highly-anticipated Meta AI. Meta has AI Studio. It's a platform for businesses to create their very own AI chatbots. These chatbots can work with Facebook, Instagram, and Messenger, making it easier for companies to connect with their customers using AI.

Get ready for an exhilarating dive into Meta’s latest hardware innovations at the Connect conference, featuring the highly-anticipated Meta AI. We’ve distilled the essential highlights for you. Keep an eye out for the freshest updates from Meta’s annual event.

Meta’s Expanding AI Chatbot Universe: From WhatsApp to Smart Glasses

Meta is making significant strides in the AI chatbot arena, introducing various AI-powered bots across its major social media platforms – WhatsApp, Instagram, and Messenger. And guess what? They’re not stopping there.

You’ll soon find these chatbots on Ray-Ban Meta smart glasses and Quest 3. These bots are powered by a ‘custom model,’ blending Meta’s Llama 2 language model with the latest in large language model (LLM) research.

But wait, there’s more! Meta has AI Studio. It’s a platform for businesses to create their very own AI chatbots. These chatbots can work with Facebook, Instagram, and Messenger, making it easier for companies to connect with their customers using AI.

Ultimate Multitasking Chatbot with Real-time Updates and Image Generation

Meta AI can assist in a range of tasks. It’s ready to help with anything from planning group trips to dropping knowledge bombs. And even if you need the latest info ASAP, it’s got real-time internet access, thanks to its partnership with Microsoft Bing. On top of that, Meta AI can create images just like OpenAI’s DALL-E 2 and Adobe’s Firefly.

Emu: Transforming Meta Chats with Innovative AI Stickers

Meta’s latest image generation model, Emu, brings exciting innovations, including generative AI stickers, soon to grace Meta’s messaging platforms. This feature empowers users to craft personalized AI stickers swiftly, spanning Meta’s suite of apps like WhatsApp, Messenger, Instagram, and Facebook Stories.

With Emu’s capabilities, users can easily create unique stickers, adding a touch of fun and creativity to their conversations across various Meta apps.

Meta’s AI Character Revolution

Meta has assembled an impressive lineup of 28 AI characters, each embodying distinct personalities, opinions, and interests modeled after cultural icons and influencers. Think Snoop Dogg and Tom Brady, but entirely AI-generated.

These virtual personalities cover everything from sports to music and social media. They are accessible for messaging on Instagram and WhatsApp.

But wait, there’s more excitement on the horizon! Initially launched in the US, these character-based AI bots are gearing up to introduce famous faces, including Bear Grylls, Chloe Kim, and Josh Richard.

Upcoming Improvements in Meta’s AI Knowledge

Meta’s AIs are still learning the ropes. Currently, its AIs are in the early stages of development. Except for Meta AI, Bru, and Perry, their knowledge is primarily derived from information available up to 2023. This means they might not have the latest updates.

However, Meta plans to boost their search capabilities in the coming months, just like it did with Meta AI. So, soon enough, you’ll get even more timely and up-to-date responses in your conversations.

Conclusion

As Meta unveils its AI-powered chatbots across platforms, it seems like a big step in improving user experiences. These innovations promise real-time information access and image generation. Emu’s image generation and the 28 AI characters add creative flair, shaping a promising AI landscape.

ChatGPT Voice AI Assistant with New Image Features by OpenAI

Posted on September 27, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence

ChatGPT now supports voice interactions, allowing users to have back-and-forth conversations with their AI assistant. Using this new feature, you can request stories, settle debates, and engage in interactive conversations with ChatGPT.

OpenAI is introducing new voice and image capabilities to ChatGPT, offering more intuitive interactions. Now, you can have more intuitive interactions with your AI assistant.

Want to have a conversation using your voice? No problem.

Need to show ChatGPT an image to discuss it? You got it!

This article explores how voice AI for business works, image features, and its potential applications in AI conversations.

Voice Conversations with ChatGPT

Exciting news! Now, you can actually talk to ChatGPT and have a back-and-forth conversation.

ChatGPT now supports voice interactions, allowing users to have back-and-forth conversations with their AI assistant. Using this new feature, you can request stories, settle debates, and engage in interactive conversations with ChatGPT. The voice capability utilizes a text-to-speech model for generating human-like audio.

But there’s more! You’re not limited to just one voice. Instead, you get to pick from five different voices to make your conversations even more enjoyable.

ChatGPT’s Voice AI and Image Understanding

Now, you can show ChatGPT what you’re talking about by sharing images! You can now share images with ChatGPT for discussions, troubleshooting, or analysis. Whether it’s fixing your grill, deciding what to cook from your fridge, or interpreting complex graphs for work, ChatGPT can provide insights based on the images you share.

Thanks to the power of multimodal GPT-3.5 and GPT-4 models, it uses language reasoning skills to understand and discuss a wide range of images, be it photos, screenshots, or documents.

Gradual Deployment for Safety

OpenAI’s strategy is all about taking things step by step to keep things safe and responsible. While voice technology is excellent, it comes with risks like impersonation or fraud. So, OpenAI is being cautious by rolling it out for voice chat first. They’ve teamed up with voice actors and partners, like Spotify, to ensure it’s used for specific, carefully considered cases, like Voice Translation.

When it comes to vision-based models for images, there are some pretty unique challenges on the table. One big concern is privacy – you definitely don’t want AI analyzing and making statements about individuals without their consent. OpenAI gets this and has taken measures to ensure ChatGPT respects people’s privacy.

Plus, they’re keeping an ear out for feedback and real-world usage to improve these safety measures. So, privacy is a top priority for them.

Transparency and Model Limitations

OpenAI believes in being transparent about what ChatGPT can and cannot do. It’s excellent at transcribing English text, but it might not perform well for some other languages, especially those with non-Roman scripts. So, if you’re using ChatGPT for specialized topics or languages, it’s less proficient in, double-checking and verifying the results is a good idea. You should use the tool wisely and understand its strengths and limitations.

Expanding Access

The great voice and image features are making their debut for Plus and Enterprise users. They get their first taste! For developers, these fantastic capabilities will soon be on the way for everyone else.

OpenAI has just significantly upgraded ChatGPT by adding voice and image capabilities. This means you can have more versatile interactions and do a whole lot more with this AI for business. It’s making your daily interactions with technology more innovative and user-friendly.

Conclusion

OpenAI’s new voice and image capabilities in ChatGPT significantly enhance user interactions with AI assistants. You can now engage in voice conversations and share images, making tasks more intuitive. Safety and privacy are paramount, with voice technology rolled out carefully and privacy measures in place for image discussions.

Transparent about its limitations, ChatGPT is a powerful tool best suited for English text. Initially available to Plus and Enterprise users, these capabilities promise to make AI interactions more innovative and user-friendly.