Storytelling with AI (Part II)

Part I explained some examples of the AI tools I use to create stories and accompanying art. It’s a disjointed process, as neither category of these AI tools can do both (yet). For text, you need to use a language model-based chatbot and for images, an art-generator app.

Language model-based chatbots are unique because they can write computer code, movie/video scripts, essays, blog and social media posts – and much more.

As a writer, I immediately saw the benefits of AI in my craft. For example, there was a short story I’d been working on (and then set aside) for years. I just couldn’t come up with an ending. I copied and pasted a summarized version into ChatGPT and asked it to “finish the story.” It did! And I was impressed.

The user interface of every AI generator I’ve used to date is rather underwhelming: A narrow Google-like text box. This suggests that concise prompts are encouraged – but I’ve used as many as 1,500 words for both ChatGPT (mainly for testing purposes). Hoolock limits entries to 1,000 characters, which is about 250 words.

ChatGPT

Hoolock

ChatGPT and Hoolock offer some suggestions to get started if you are new to these AI tools. But I use them a bit differently.

After work (and a cold beverage), I sometimes get an idea for a short story in my head that – without AI – I’d jot down for later consideration. But with AI at my fingertips, would it kill me to spend 30 seconds entering the idea into one of my two new creative tools? No!

This morning I wrote the following introduction to a fictional account of a fishing boat that was being overtaken by a storm:

The small fishing boat bobbed violently as the choppy waves kicked up by the hurricane continued to pummel the ship. Counter to his best efforts to outrun the tempest, the mighty storm was overtaking the two-man craft.

“Secure the nets, Mr. Schmitt!” Capt. Pauly shouted to his first mate. “We’ve no choice but to ride this out to the end!”

I titled it “Capt. Pauly’s Plight.” This is what a plugged into ChatGPT:

If you want to know how this tale ends, visit my blog – ChatGPT picked up where I left off here!

I then used what I had written along with a paragraph or two that ChatGPT contributed and pasted that text into Hoolock. Here’s the art that popped out the other end:

Hoolock-generated image of a fisherman on a boat

While I donated this piece to support “disaster relief,” I took a screen grab of just the image to use in my blog post. I took the liberty of doing so since I’m the author – but soon, the founders of Hoolock will make it possible to link the image back to their site for purchase considerations. It’s on the roadmap, they told me!

Storytelling with AI (Part I)

As a writer, I immediately saw the benefits of AI in my craft: for words, language model-based chatbots and for images, art-generator apps.

For words: Language model-based AI chatbots can carry out human-like conversations and write natural, fluid prose on various topics. They can also perform complex tasks, from writing programming code to planning your daughter’s wedding.

For art: An AI image generator uses machine learning called artificial neural networks that take your text input and turn it into an image.

I use both models for specific work-related projects: creating video scripts, creating succinct summaries of messy meeting transcripts, creating social media posts from blog posts, news releases and announcements – and many more.

Both AI tools have variations, but my favorites are ChatGPT (for the words) and Hoolock (for the art). There are paid options for some AI tools, but I’m happy with the free versions for now.

I like ChatGPT because, in my experience, it delivers superior results than the competition. And for art, I use Hoolock – not only because I like the results but also because it has brought out the “inner AI artist” in me.

Best of all, with Hoolock, I’m turning my words into very cool pieces of digital art that I can donate to a non-profit that can then sell the art to raise funds for the cause it represents.

How is that possible? Isn’t digital art just an image anyone can download? Yes, but in this case, that specific piece of purchased art resides on a downloadable non-fungible token (NFT) stored on a blockchain. A blockchain is a decentralized, distributed, public digital ledger that cannot be hacked. Think of a blockchain as the digital version of the bookkeeper’s infamous ledger in the movie “The Untouchables.”

Think of a blockchain as the digital version of the bookkeeper's infamous ledger in the movie “The Untouchables.”

Only the owner of the NFT can download it. Afterward, the NFT can be used on mugs, t-shirts, books, websites – anywhere and anyhow the owner wants to use it.

Read Part II of this post to learn some techniques I use to create AI-generated fiction and supporting artwork.

Empowering AI Clusters: Cisco’s Ethernet Technology Leads the Way

The rapid advancements in Artificial Intelligence (AI) have revolutionized industries across the globe, from healthcare to finance and manufacturing to transportation. However, the sheer complexity of AI workloads demands robust computing power, leading to AI Clusters’ emergence.

In this blog post, we will explore what Ethernet Technology is and how Cisco’s cutting-edge solutions are elevating the performance and resiliency of AI Clusters, propelling AI research and development to new heights.

Understanding AI Clusters

AI Clusters are high-performance computing systems specifically designed to handle the enormous computational requirements of AI applications. These clusters consist of interconnected computers or servers working collaboratively to execute intricate AI algorithms at incredible speeds. Through parallel processing and distributed computing, AI Clusters enable researchers and data scientists to analyze vast datasets, train sophisticated machine learning models, and unlock the true potential of AI.

The Significance of Ethernet Technology

Ethernet Technology, the backbone of modern networking, is a standardized family of technologies for data transmission over local area networks (LANs) and wide area networks (WANs). It provides the foundation for seamless and reliable communication between devices, facilitating data transfer, audio, and video across diverse IT environments. Cisco, a leading networking technology provider, has been at the forefront of developing and innovating Ethernet solutions to the ever-evolving needs of modern computing infrastructures.

Enhancing the Performance of AI Clusters

  1. High-Speed Data Transfer: As AI workloads demand the swift movement of massive datasets, Ethernet’s high-speed capabilities play a pivotal role in reducing data transfer bottlenecks. Cisco’s Ethernet solutions, such as 25 Gigabit Ethernet (GbE), 40GbE, and even 100GbE, ensure that AI Clusters can move and process data without compromising performance.
  2. Low Latency: Real-time AI applications, such as autonomous vehicles and natural language processing systems, heavily rely on low-latency networking. Cisco’s advancements in Ethernet technology provide ultra-low latency communication, reducing the time required for data to travel between nodes in the cluster. This optimized data flow enables AI Clusters to deliver faster and more accurate results.
  3. Scalability: AI research is an iterative process that requires constant experimentation and adaptation. Cisco’s Ethernet technology allows for seamless scalability, making it easier to expand AI Clusters as computational demands increase. This flexibility ensures that AI researchers can adapt their infrastructure to tackle complex AI tasks without network capacity limitations.

Boosting Resiliency and Reliability

  1. Network Redundancy: AI Clusters require a robust and reliable network to prevent data loss and ensure smooth operation. Cisco’s Ethernet solutions incorporate redundancy protocols, such as Rapid Spanning Tree Protocol (RSTP) and Link Aggregation Control Protocol (LACP), to protect failover and ensure uninterrupted data flow even during network failures.
  2. Traffic Prioritization: With AI Clusters often hosting multiple concurrent applications, proper traffic prioritization is crucial. Cisco’s Ethernet technology supports Quality of Service (QoS) mechanisms, enabling administrators to allocate bandwidth based on application priorities. This feature ensures critical AI workloads receive the necessary network resources, optimizing cluster performance.

The future of AI research and development is intrinsically tied to the performance and resiliency of AI Clusters. As the demand for computing power continues to rise, the role of Ethernet Technology in enhancing these clusters cannot be overstated. Cisco’s dedication to pushing the boundaries of networking technology has resulted in cutting-edge Ethernet solutions that empower AI Clusters with unmatched speed, reliability, and scalability. Through their innovative contributions, Cisco is undoubtedly playing a vital role in accelerating AI advancements and opening new doors to unprecedented possibilities across various industries.