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.

What Does Amazon’s Latest Generative AI Tool for Sellers Offer?

Posted on October 17, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence, Business, Business Growth

Amazon’s latest move in the e-commerce game is its generative AI for sellers, launched at its annual seller conference, Amazon Accelerate 2023. Thanks to new AI capabilities, creating compelling, useful product listings just got a whole lot simpler! In this blog, we’ll find out what it is all about.

Amazon’s Generative AI for Sellers

Amazon has stepped up its selling game by bringing generative AI for sellers into the mix. Thanks to its newly introduced set of AI capabilities, Amazon sellers can create detailed and engaging product descriptions, titles, and listing details more easily.

Yes, that’s right! No long, complicated processes. Sellers won’t need to fill out lots of different info for each product anymore. It’ll be much quicker and simpler to add new products. This way, they can enhance their current listings, giving buyers more assurance when making purchases.

Source

“With our new generative AI models, we can infer, improve, and enrich product knowledge at an unprecedented scale and with dramatic improvement in quality, performance, and efficiency. Our models learn to infer product information through the diverse sources of information, latent knowledge, and logical reasoning that they learn. For example, they can infer a table is round if specifications list a diameter or infer the collar style of a shirt from its image,” shares Robert Tekiela, vice president of Amazon Selection and Catalog Systems

What Exactly Does Amazon’s Generative AI for Sellers Do?

Here’s what Amazon’s new AI capabilities bring in for sellers:

Sellers only need to submit a brief summary of the item in a few words or sentences, and Amazon will create high-quality text for their review.
If they want, sellers can edit them.
They can just submit the automatically created content to Amazon’s catalog.

The result? High-quality listings for sellers. And guess what? Shoppers will have a better time finding the product they want to buy.

How Does Amazon AI for Sellers Work?

Amazon has used machine learning and deep learning to automatically extract and improve product information. More specifically, it uses large language models (LLMs) to create more thorough product descriptions. But why LLMs? Well, these machine learning models are trained on vast volumes of data. So, they can detect, summarize, translate, predict, and generate text and other material.

Note that the American e-commerce giant didn’t say exactly what information it used to teach its LLMs. However, it appears that the company might be using its own listing data.

The use of generative AI models at such a large scale, however, raises certain concerns: the propensity to generate untrue, erroneous information, plus other errors that may go unnoticed unless a human checks them.

Nonetheless, over the past few months, many sellers have tested Amazon’s newest AI products, and preliminary feedback suggests that the majority of them are actively using the AI-generated listing content for them.

Conclusion

Amazon is starting to make it simpler for listing creators to use AI, which is just one of the ways it is assisting sellers in starting and growing profitable businesses. This is only the start of how it intends to employ AI to enhance the seller experience and support more successful sellers.

Read More: AI Studio by Meta

What is Mistral AI: Open Source Models

Posted on October 14, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence

The French startup Mistral AI has introduced the GenAI model. Is it the next best AI business assistant?

In a big step to disrupt the AI field, the French startup Mistral AI has introduced the GenAI business assistant. It’s ready to take the dominance of industry giants like Meta and OpenAI. This blog explores the potential implications of this exciting development in artificial intelligence.

Mistral AI’s Astonishing $113 Million Valuation: What’s the Buzz?

Mistral AI, a Paris-based AI start-up, grabbed a lot of eyeballs when it raised a huge $113 Million at a valuation of a massive $260 Million. The company was only three months old and had less than 20 employees. So, it seemed like a valuation game at that time.

Fast forward a couple of months, Mistral AI has launched its own open-source large language model Mistral 7B. It is better across all parameters than the Llama 2 13B model, which is twice as big as Mistral 7B. Mistral AI is also better than Llama-1 34B on many benchmarks.

Mistral 7B vs. the Giants: How This AI Open Source Outperforms

This lightweight AI model is competing with existing heavy-weight AI models. And it isn’t backing out!

The performance of Mistral AI so far, at a fraction of the cost and resources, has proved that it is worthy of its huge valuation. Here are some of the major reasons for Mistral AI’s success:

Training methods used by Mistral AI to train its first-gen AI model are more efficient.
Mistral AI’s training methods are at least two times less expensive to implement than the existing methods.
The open-source nature provides greater flexibility.
The open-source model is easy to fine-tune, which is the cherry on the top.

Mistral AI has made these models open to all. So, does that mean this French start-up will be coming up with bigger, better, and more complex models? Well, yes!

Up until now, AI enthusiasts around the world have been dependent on Meta to come up with good-quality AI business assistants and foundation models. So, Mistral AI’s GenAI Model is a good thing that happened to them.

Paving the Way for New AI Players

The AI assistant sector has been an oligopoly, with a majority of its players being from the US. But what’s been keeping other players at bay till now? The reason is the high barrier to entry. It requires hard-to-make technology and a tremendous amount of investment to compete with these potential AI employee giants.

With millions of dollars in funding and the rarest of the rare team, Mistral’s entry can cause disruption in this field. In fact, Mistral is looking to develop an AI assistant for business superior to GPT-4 as soon as 2024, just like LLaVA.

What sets Mistral apart in the AI field? The founding team of Mistral consists of leaders in the field of AI assistants for business. With experienced researchers, formerly from Meta and DeepMind, Mistral’s fast-paced success is no fluke, and their future plans to rival Meta and OpenAI seem well-thought-out.

The flexibility and open-source license of Mistral AI’s new AI business assistant model provide an even ground for everyone to enter the AI space. However, since this model can be used without restrictions, its ethical use could be a matter of concern.

Conclusion

Mistral is riding the AI wave smoothly, and this French start-up is ready to give tough competition to proprietary AI solutions for business provided by Meta and OpenAI, all within a couple of years since its inception.

Now that there is another big player in the scene, you can expect to see other kinds of models as well, not just language models. Such high-quality open-source models show a shift in the AI industry. It signifies that new business AI models like Mistral AI are here to directly compete with US AI giants like Meta and OpenAI.

Read More: Top 6 AI Tool Directories in 2023

Meta’s AI Studio: Create Your Own AI Chatbot, Tool, and Software

Posted on October 11, 2023 by Oriol Zertuche - AI tools, Artificial Intelligence, Business

At the recent Meta Connect 2023 event, Meta CEO Mark Zuckerberg introduced a range of AI experiences for individuals and businesses, including AI Studio. Using AI Studio, you can create your own AI chatbot, tool, or software! With 1.5 billion AI chatbot users worldwide, Facebook’s parent company, Meta, aims to make AI development available to everyone.

Meta’s new AI innovation gives you the power to make personalized AI chatbots without any coding expertise.

“There is clearly the small business and enterprise side of this as well, primarily in terms of productivity, better communication and user engagement,” says Arun Chandrasekaran, an analyst at Gartner.

Thanks to a span of pre-trained models and user-friendly drag-and-drop tools it offers, AI Studio lets anyone craft and train their AI chatbots. From customer service chatbots to AI chatbots that talk like celebrities or historical personalities, AI Studio’s creative potential has no bounds!

Meta Contributing to the AI Ecosystem

From Generative AI and Natural Language Processing (NLP) to Computer Vision and other core areas of AI, Meta has long focused on connecting people in fun and inspiring ways through collaborative and ethical AI solutions. Meta Connect 2023 also witnessed the launch of AI stickers, Emu for image editing, Ray Ban smart classes, Quest 3, and more.

Watch! Origin Stories – Meta AI

In 2016, Meta, then called Facebook, released a Messenger development kit for messaging chatbots geared toward businesses. This is when AI Studio was first introduced. But fast forward to today, and these AI Studio bots are nothing like the rigidly programmed rules-based bots of the past. They are more capable and dynamic in their answers.

How?

Well, they’ve been using powerful language models.

One of them is Meta’s Llama 2, trained on more than 1 million human annotations.

And guess what’s happening in the coming weeks? Developers can use Meta’s APIs to create third-party AIs for its messaging services. This development will kick off with Messenger. Instagram and WhatsApp are next in line.

From small businesses aiming to scale to huge brands wanting to improve communications, every firm will be able to develop AIs that enhance customer service and embody the values of their brands. AI Studio’s main use case right now is E-commerce and customer support. Although Meta has started with an alpha version, it plans to expand and refine AI Studio in 2024.

On top of it, creators will be able to develop AIs that spice up their digital presence across all of Meta’s apps. They’ll be able to approve these AIs and have direct control over them.

Meta’s AI Sandbox and Metaverse Synergy

Alongside the debut of AI Studio, Meta spilled the beans about a sandbox tool coming your way in 2024. This platform will allow users to play around with AI creation, potentially democratizing the creation of AI-powered products.

What’s even more amazing? Meta has big plans for integrating this sandbox tool into its metaverse platforms. One such platform is Horizon Worlds. This will let you enhance a variety of metaverse games and experiences made using AI Studio.

Conclusion

With AI Studio’s advanced capabilities addressing a range of chatbot requirements, coupled with the sandbox tool, Meta’s efforts toward making AI accessible for all can be expected to transform the AI chatbot arena for professional and personal usage.

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.