With the latest AI darling, DeepSeek AI, wiping billions off the market value of US tech giants just yesterday, 2025 is already shaping up to be a fascinating year for AI. The rapid evolution of AI, its promises, pitfalls, and shifting priorities, sets the stage for a year full of disruption. Here are my predictions for what’s IN and what’s OUT in AI for 2025:
AI Tech Stack: OUT with Training Obsession, IN with Inference*
The obsession with training massive models is OUT. What’s IN? Ruthlessly efficient inference. In 2025, if you’re not optimizing for inference, you’re already behind. Here’s why.
The cost of achieving OpenAIo1 level intelligence fell 27x in just the last 3 months, as my Google Cloud colleague Antonio Gulli observed – impressive price-performance improvement.
The recent DeepSeek AI breakthrough proves this point perfectly. Their R1 model (trained for just $5.6 million, a fraction OpenAI’s rumored 500 million budget for its o1 model), achieves feature-parity and even outperforms major competitors in key benchmarks:
We clearly figured out how to make LLM training more effective and cost efficient. Time to reap the benefits and use the models for inference.
*We will still be enhancing LLMs’ capabilities, developing smaller, purpose-built models and re-training them with new data-sets.
AI Architecture: OUT with Cloud-First, IN with Edge-First**
The pioneers in the most AI-advanced industries like Manufacturing have exposed the limitations of cloud-first AI approaches. According to Gartner, 27% of manufacturing enterprises have already deployed edge computing, and 64% plan to have it deployed by the end of 2027. Why the rush to edge-first AI architectures?
In industrial applications, especially those requiring real-time control and automation, latency requirements as low as 1-10 milliseconds demand a fundamental rethinking of distributed AI system design. At these speeds, edge-to-cloud roundtrips are impractical; systems must operate as edge-native, with processing and decision-making happening locally at the edge.
One of Synadia‘s most innovative customers, Intelecy, a No-Code AI platform that helps industrial companies optimize factory and plant processes with real-time machine learning insights, perfectly illustrates this paradigm shift. Their initial cloud-first approach had processing delays of 15-30 minutes. By redesigning their AI architecture for the edge, they achieved less than one-second round-trip latencies. This dramatic improvement enabled real-world applications like automated temperature control in dairy production, where ML models can provide real-time insights for process optimization.
Processing data where it is generated isn’t just more efficient—it’s becoming a competitive necessity for every industry. Gartner predicts that by 2029, 50% of enterprises will use edge computing, up from just 20% in 2024.
**The cloud’s role in AI isn’t disappearing (of course), but the default is shifting rapidly towards edge-first thinking.
AI Impact: OUT with What-If, IN with What-Now***
Focusing on model capabilities is OUT. What’s IN? Solving real business problems. The most compelling AI stories in 2025 won’t mention model architecture. Instead, they’ll focus on measurable business impact.
Intelecy’s Chief Security Officer 🔐 Jonathan Camp explains how AI can help ensure quality in manufacturing: “A dairy can use a machine learning forecast model to set temperature control systems using the real-time predicted state of the cheese production process.The process engineering team can use Intelecy insights to identify trends and then automate temperature adjustments on a vat of yogurt to ensure quality and output are not compromised.”
The shift is clear: success is no longer measured in model capabilities, but in hard metrics like revenue gained, costs saved, and efficiency improved. The question isn’t “What can AI do?” but “What value did it deliver this quarter?”
***As an innovation-obsessed marketer, I’l never give up on “what-if” dreams but “what-now” is the state of AI in 2025.
The Elephant in the Room: Can gen AI be trusted?
We’ve solved training costs. We’ve started to crack real-time processing. Now, the focus shifts to trust: Can AI deliver consistent, reliable, and verifiable results at scale?
For example, try to ask 3x gen AI bots this prompt 3x, and see for yourself:
Name top 3 ski resorts in Europe by the total length of ski runs that are truly interconnected (no bus transfers)
We’re entering the era of agentic AI where AI-made decisions will be automatically implemented by chains if AI-functions. Are we ready?
As a technology marketer, I’ve witnessed the Gen AI revolution from its inception. In early 2023, I crafted an enterprise narrative for Google Cloud, helping our global salesforce inspire customers to adopt this technology. I’ve seen AI evolve significantly: from chatbots writing silly poems to answering medical questions and guiding students in physics. Today, AI can understand, reason, and create across various inputs like text, images, audio, and video. I’m excited about AI’s potential to accelerate marketing innovation.
Now, as the VP of Marketing at Synadia, the startup on a mission to connect the world, I’ve observed even more. At a recent webinar for 50+ portfolio companies at Forgepoint Capital, I shared these insights, which I’m highlighting in this week’s newsletter. Thank you Tanya Loh for the opportunity!
As a technology marketer, I’ve witnessed the Gen AI revolution from its inception. In early 2023, I crafted an enterprise narrative for Google Cloud, helping our global salesforce inspire customers to adopt this technology. I’ve seen AI evolve significantly: from chatbots writing silly poems to answering medical questions and guiding students in physics. Today, AI can understand, reason, and create across various inputs like text, images, audio, and video. I’m excited about AI’s potential to accelerate marketing innovation.
18 months ago, I visualized gen AI as my alter ego, a game changer amplifying our strengths and overcoming our weaknesses. I hoped that for creative and curious people who tend to procrastinate and get bored easily, Gen AI would make life easier by handling the repetitive and mundane tasks we dislike.
Gen AI amplifies many our strengths and overcomes our weaknesses
Today, I can say that my prediction turned out to be true. I see three main buckets where Gen AI helps us in marketing:
✅ Expand your expertise
✅ Jump-start your creativity
✅ Offload your marketing tasks
I’ll illustrate how I’ve used Gen AI in marketing through stories with Google and Synadia.
Gen AI quickly expands our expertise and teaches us new skills
First, let me tell you how Gen AI helped me hit the ground running from Day 1 at Synadia and how it continues to my invaluable sidekick. I created a custom GPT to teach me about Synadia’s product portfolio. I trained it using public documentation. I knew the learning curve on the product built by the world’s brightest distributed systems engineers would be steep. Asking my personal GPT questions promptly got me up to speed.
Here’s another example. Imagine hearing a new acronym during a meeting with your engineers. Instead of interrupting the flow and making everyone wait while someone explains it, my Synadia bot, trained on github.com/synadia-io, immediately clarifies it for me. This way, we can stay focused on discussing our vision for building the ultimate platform for distributed applications without sidetracking the meeting.
My other story happened at Google. You may have seen ‘The Tale of a Model Gardener’ video, a humorous cartoon about Gen AI’s ability to help enterprises achieve their business goals. Gen AI accelerated the creative process for this video we launched ahead of #GoogleCloudNEXT in August 2023. Here’s how.
The idea hit me when cycling between San Francisco’s Marina District and Financial District to Google’s office. I wondered: “How do I explain concepts in AI, such as augmentation and prompt engineering, in fun, approachable ways?” The idea of a cartoon video surfaced but how to start, having never written a video script nor cartoon?
With just 30 minutes until my next meeting, I instructed Bard (now #GoogleGemini). “Hey, Gemini, write me a movie script about X.” I added three sentences with my idea. In minutes, I got a fully developed movie script, beautifully formatted by scenes. I perfected things with small additional prompts. Google #Vertex AI, an end-to-end ML platform, helped me generate images. I stitched together a script draft with some images and sent it to my creative colleagues–all within 30 minutes–and they really liked the idea of a cartoon. Though they had different ideas about what the cartoon ought to explain and how it ought to look; the concept landed from my Gemini experiment.
This triumph really shows what Gen AI brings us as marketers. I got a solid movie script within minutes, with no prior experience in building such things. I didn’t waste the idea, which became an important, valuable deliverable for the company.
Have you seen this fairy tale 🐇🥕🥧 about gen AI?
Gen AI sparks our creativity
I want to expand on how I use Gen AI to jumpstart my creativity, especially when short on time. As you can imagine, only a few weeks into the Synadia role, I’m working on our positioning and messaging. I brainstorm with my small but mighty marketing team, my product and engineering team, and my founder and CEO. I also brainstorm with Gen AI, especially when everyone’s busy. (My bot is always available.)
As the proud granddaughter of a professor of physics and a prolific dressmaker who whipped up gorgeous fashion from his patterns, I’ve long loved prototyping and testing my ideas. My mind works best when reacting to prototypes vs thinking about them. Prior to Gen AI, we had to write or sketch out our prototypes. Gen AI requires a simple prompt for a full document which can spark more ideas and creativity in ourselves and others. (We saw this with my design team at Google and ‘The Tale of a Modern Gardener’ AI-generated cartoon script.) For that same reason product demos are worth 1000 slides. Show, don’t tell.
Have you read our
Gen AI offloads small tasks
Gen AI also can offload small, repetitive, mundane tasks to free us up for more strategic thinking and exciting tasks. At Synadia and Google, Gen AI has helped me:
Jump-start projects. A custom prompt to generate case studies from our many great customer stories captured in blog posts and videos scaled our small team’s output.
Generate images. The early images for my first cartoon script for ‘The Tale of a Model Gardener’, weren’t perfect, but brought the narrative to life.
Edit content and minor things/ideas. While preparing my creative idea for the design team, I lacked the time for editing parallel construction in my lists or capturing typos. My bot took care of that so I could focus on creativity. Writing uses the creative part of our brain; editing uses the analytical. Mixing the two puts the breaks on the creative process so I like to offload the latter to my bot.
Gen AI helps us feel more experimental
Gen AI never lets an idea go to waste. When you’re pressed for time, quickly producing a first prototype helps your colleagues provide feedback faster. While you might discard that initial version, it speeds up the journey to the final product.
If a picture is worth a 1000 words, a prototype is worth a 1000 thoughts
Sometimes, a colleague may take your prototype in a completely new direction. I find this process empowering and encouraging. Each strong reaction, even a negative one, means I’m one step closer to the ideal solution.
Innovation thrives on collaboration and diversity, not egos. Gen AI helps create an environment where ideas evolve and improve through teamwork, making our solutions stronger and more innovative.
The drawbacks and obstacles with Gen AI
Gen AI has some drawbacks. A few I’ve encountered include:
Ubiquitous language. The bar for quality content has never been higher and savvy readers detect and tune out AI-generated if repetitive, generic, and vague. We’ve seen content overload for close to two years now. Cutting through that noise requires high-quality content.
Flawed responses. Use AI responsibly and not verbatim. AI bots are not deterministic. Our bot’s responses may be quick, but sometimes contain significant errors in reasoning. I wrote about this problem in my post on how I turned a daunting 150+ page-long voter pamphlet into a handy cheat sheet for the San Francisco elections. I prompted: “Summarize all the propositions on the March 5th 2024 SF election ballot with their top arguments for and against in a 3-column table using the voter pamphlet as the data source.” The bot’s quick response impressed me. But I found reasoning errors and one argument was entirely made up by the bot. Always check your results.
No slide fixes! Gen AI will not do what we dream of (yet): unify our fonts and texts in our slide decks 😂.
Change in Gen AI is unprecedented. I’ve seen nothing like this growth in my 15+ years in tech. The new features from OpenAI, Google or Anthropic are just the tip of AI innovation. Many startups work towards perfecting Gen AI as well. In the meantime, discovering gen AI feels amazing and I wonder how we lived without it.
Looking ahead
Despite all that Gen AI brings us as marketers, it cannot compete with human storytellers. Gen AI does not substitute well-written, well-narrated customer stories. Even OpenAl looks for interesting use cases of their products exploring new features in unexpected ways. So let’s provide them. A recent LinkedIn post on me using ChatGPT and a Peloton app to rediscover German became one OpenAI reposted, which sparked a wonderful conversation on learning new languages with Gen AI, all from a lighthearted, personal story connecting with technology, efficiency, and learning.
The previous edition of this newsletter reposted by Open AI
This moment reminds me: All Tech brands, even the Silicon Valley hottest companies like OpenAI, seek interesting stories on how we use their products in exciting, unexpected ways to start a community conversation.
For now, Gen AI cannot do that nor can it replace a great writer or story. That’s our opportunity and another way we can best partner with Gen AI as marketers and as storytellers.
➡ How do YOU use Gen AI in Marketing? Share your thoughts in the comments! ✍👇
➡ Need a hand getting started? Shoot me a message! ✍ ✉
The singer Rihanna “found love in a hopeless place,” and I found my teenage crush in the back of a Ross store in Pasadena during the July 4th weekend.
There he was: Prof. Richard Feynman, whom I first heard about in high school Physics class. His face on a giant mural, taking over the entire back wall of a warehouse-like building. ➡ To a young aspiring scientist, he was the ultimate role model: a world-famous Physicist and a Nobel Prize winner. ➡ To a marketing executive, he suddenly appeared as my first marketing icon.
Prof. Feynman could explain quantum physics in ways that excited the general public. As an #engineer-turned full-stack technology #marketer, I’ve always felt inspired to do the same. If Feynman could make physics interesting and relatable to everyone—even teens—then I must strive to explain any technology I’m responsible for marketing in clear, layperson terms.
Feynman’s “teaching method” has become my gold standard. I often ask myself: “How would Feynman explain it?” With human attention spans dropping to less than that of a goldfish 🐠, the pressure to communicate clearly is immense. (Did you know that goldfish have an average attention span of 9 seconds, while humans have a 8.25 second attention span?) https://lnkd.in/gxGqMsqh
Just a few weeks into my new role as VP of Marketing at Synadia, I’m inspired to channel that same spirit of simplicity and excitement into my work. For example, here’s how we make complex #Microservices systems simple with NATS.iohttps://lnkd.in/gd-BZZMN: 3 favorite things in a technical post: ✅ architecture diagrams ✅ call flow diagrams 🍒 ant lines in the diagrams 🐜 as a cherry on top!
Just like the mural in Pasadena blends art and science in unexpected ways, so should marketing in 2024. We must cut through the noise and capture our audience’s attention in those precious few seconds ⏱️.
What techniques do you use to make complex ideas relatable to your audience?👇
“The Magic of Generative AI” is still my favorite talk I’ve ever given, hands down. I loved collaborating with Google’s top AI minds on the story and the visuals, building demos that showed how Vertex AI helps marketers like me, and connecting with fellow AI enthusiasts in awesome places like LA and Rome.
But the best part was diving deep into how large language models (LLMs) actually work, reading those mind-bending research papers, and piecing together the “magic” they create. Preparing this talk was like living Google’s innovation mantra: stay curious, experiment, build something useful.
In this newsletter, I’m sharing my reflections on the magic of Gen AI and how Google’s unique innovation culture was key to making these incredible tools a reality.
Curiosity, experimentation, and application: This is the heart of how Google is driving the generative AI revolution. It’s the same formula behind some of our biggest breakthroughs, like Google Search, Translate, and Vertex AI.
Here’s how it works:
Curiosity: This is where it all starts – that burning question of “what if?” or “why not?” Curiosity is what drives us to explore the unknown and challenge the status quo.
Experimentation: Curiosity without action is just daydreaming. Experimentation is where we get our hands dirty, trying new things, making mistakes, and learning from them. It’s the messy but essential part of the process.
Application: The ultimate goal of innovation is to create something that makes a real difference in the world. Application takes those wild ideas and experiments and turns them into practical solutions that people can use and benefit from.
This isn’t just a theory; it’s the blueprint behind Google’s most groundbreaking AI tools.
Embeddings in Google Search: Grasp query intent beyond exact keywords
In 2013, Google researchers authored the seminal paper “Efficient Estimation of Word Representations in Vector Space“. This paper unveiled a revolutionary method for creating Word Embeddings, mathematical representations of words capturing both their meaning (semantics) and relationships (semantic similarity). Here’s the Google’s innovation formula in action:
Curiosity: Dissatisfied with existing word organizational methods, such as dictionaries ordering words by lexicographical order, researchers were curious if a better approach could capture word semantics and organize them by semantic meaning.
Experimentation: They explored various neural network types, training objectives and relationship representations. Through experimentation, they discovered how to automatically create a word embedding. A name to be remembered, an embedding is a mathematical representation for each word that captures their semantic meaning in the form of a vector of 768 numbers.
Application: Way before it was applied in gen AI, word embeddings found a magical application in semantic search, enabling Google Search 🔍 to grasp query intent beyond exact keywords. For example, a search for “cars that are good on gas” now returns results for fuel-efficient cars, even if the word “gas” doesn’t appear in the options returned.
Source: “The Magic of Generative AI” talk, Google Gen AI Live and Labs event series
Transformer in Google Translate: More accurate translations
In 2017, Google researchers presented “Attention is All You Need” introducing the Transformer architecture, built on decision-making and attention-span concepts. It empowers the language models to understand context and relationships within word sequences. Curiosity, experimentation and application were again vital:
Curiosity: In the search to improve the quality of language translation, researchers sought ways to model relationships among words in a sentence.
Experimentation: They experimented with various mechanisms, relationship representations and training methods, discovering that much could be extracted by simply paying attention to the relationship between each word and every other word in a sentence. They discovered that these interdependencies could be achieved through parallel computations, which accelerated time to result, and found that representations through embeddings could capture long-range dependencies between words in fluent, grammatically correct text. Voilà! The Transformer architecture was born, introducing a huge breakthrough in science.
Application: The transformer revolutionized Google Translate 🌐. The Transformer’s attention mechanisms are excellent at understanding the relationships between words in a sentence, leading to more accurate translations.
Source: Transformers, FT,
Let’s see this in action by translating this sentence from English to Italian: “The cat didn’t cross the street because it was too wide.
Source: “The Magic of Generative AI” talk, Google Gen AI Live and Labs event series
Gen AI in Enterprise Search: New way of working
Fast forward to 2023, Google Cloud researchers set to simultaneously tackle two common challenges for many organizations:
How to organize enterprises information scattered across many internal systems
How to make this information accessible and useful for enterprises, and seamlessly available and actionable in applications such as customer service bots, document summarizations or as part of steps in automated workflows.
Not surprisingly, Google Cloud researchers followed the proven innovation framework:
Curiosity: While Google Search was designed to scale to organize the world’s information, researchers started exploring whether the technology could be descaled and made available to enterprises to organize their information in a way that could be easily accessible and useful to them, and only to them.
Experimentation: Intrigued by the potential to bring together several cutting-edge technologies, researchers used the ability to crawl web-sites to discover content on internal websites and structured content, and Optical Character Recognition (OCR) to discover content from all sorts of semi-structured and unstructured documents, creating a wealth of knowledge about the enterprise. The researchers then used embeddings to extract and organize the semantic meaning of all of this data. Once enterprise’s data has been semantically organized in embeddings, the full power of Generative AI can be applied to it and leveraged across the Vertex AI platform.
Application: First launched in March 2023, Google Cloud Vertex AI Search 🔍 quickly became “the killer enterprise app”. A killer application, often abbreviated as killer app, is a software application that is so necessary or desirable that it proves the core value of some larger technology, such as its video game console, software platform, or in this case of gen AI in the enterprise context. Killer apps are the pinnacle of innovation: well-designed, easy to use and solving a real problem for users. Enterprise Search is the killer enterprise app as it unlocks unprecedented levels of productivity and efficiency.
These transformative breakthroughs exemplify Google’s dedication to AI innovation, with continued explorations on the horizon.
With the entire text of Les Misérables in the prompt (1382 pages), Gemini 1.5 Pro locates a famous scene from a hand-drawn sketch
Technical writing is one of my favorite reads. It’s clear, succinct, and informative. DeepMind’s technical paper on Gemini 1.5 epitomizes all I love about technical writing. Read the abstract for a glimpse into the groundbreaking advancements encapsulated in Gemini 1.5 Pro; it’s a masterclass in effective communications. We learn how to deliver maximum insight with minimum word count.
In just 177 words, my DeepMind colleagues articulate:
#ProductCapabilities: “a highly compute-efficient multimodal* mixture-of-experts model** capable of recalling and reasoning*** over fine-grained information from millions of tokens of context”
#UniqueSellingPoint: “near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 2.1 (200k) and GPT-4 Turbo (128k)”
#UseCases: “surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person learning from the same content”
Gemini 1.5 Pro is able to translate from English to Kalamang with similar quality to a human
The science of writing succinctly
In a few words, the paper abstract communicates the model’s superior performance, its leap over existing benchmarks, and its novel capabilities. It sparks curiosity about the future potentials of large language models—a true testament of powerful, precise, impactful technical communication.
How did the Gemini 1.5 paper authors achieve this mastery? By following the guiding principles of Brevity (saying more with fewer words) that my friend and thought partner D G McCullough and I recently summarized as: “Trust, Commit, Distill”:
#Trust means believing in the power of your message without over-explaining nor adding unnecessary details. Trust empowers the communicator to eliminate redundancy, focusing on what’s truly important. The Gemini 1.5 paper authors trust their curious readers to look up terms that may be new to them. On first read, I had to look up “mixture-of-experts” but the context I’ve had from my 2 years of working with data and AI allowed me to “guesstimate” its meaning before getting the proper definition.
#Commit refers to sticking with the essentials of your message, understanding your message’s objective, and resisting tangents or unnecessary explanations diluting the message’s impact. (Which requires discipline!)
#Distill requires breaking down your message to full potency. Like distilling a liquid to increase its purity, we must strip away the non-essential until the most impactful, clear, and concise message remains. Every word and idea then serves a purpose–and voila! Your message becomes clearer, and more memorable.
The art of replacing 100s of words with a single image
The saying “A picture is worth a thousand words” truly shines in technical communication. A single, well-chosen image can articulate complex ideas with more efficiency and impact than verbose descriptions. The Gemini 1.5 paper’s authors skillfully weave in visual elements, showcasing a deep grasp of conciseness. This approach not only makes complex AI and machine learning concepts approachable and captivating but also boosts understanding and enhances the reader’s journey. It demonstrates that when it comes to sharing the latest scientific breakthroughs, visual simplicity can convey a wealth of information.
With the entire text of Les Misérables in the prompt (1382 pages), Gemini 1.5 Pro locates a famous scene from a hand-drawn sketch
Simplify complexity with brevity
In our rapid world, where attention is a rare commodity and people often skim rather than read, the skill of conveying ideas briefly and through visual storytelling stands out as a significant edge. Simplifying complex concepts into engaging visuals and concise explanations can mean the difference between being noticed or ignored.
Richard Feynman, the celebrated physicist, Nobel laureate, and cherished educator, famously stated, “If you can’t explain it simply, you don’t understand it well enough.”
Richard Feynman quotes
Feynman’s approach isn’t just about words; it involves using visuals and images to make intricate ideas more approachable. After all, the deepest insights are usually the easiest to understand when we apply brevity to break down complexity.
DeepMind’s Gemini 1.5 technical paper exemplifies this principle perfectly. It’s essential reading for anyone intrigued by general AI (especially with #GoogleCloud#NEXT24 on the horizon), and it’s an exemplary model for those dedicated to honing their communication skills.
“In this report, we present the latest model of the Gemini family, Gemini 1.5 Pro, a highly compute-efficient multimodal mixture-of-experts model capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. Gemini 1.5 Pro achieves near-perfect recall on long-context retrieval tasks across modalities, improves the state-of-the-art in long-document QA, long-video QA and long-context ASR, and matches or surpasses Gemini 1.0 Ultra’s state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5 Pro’s long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 2.1 (200k) and GPT-4 Turbo (128k). Finally, we highlight surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person learning from the same content.” https://storage.googleapis.com/deepmindmedia/gemini/gemini_v1_5_report.pdf
Define the key terms used in the abstract
* #Multimodality: Gemini is natively multimodal. Prior to Gemini, AI models were first trained on a single modality, such as text, or image, and then corresponding embeddings were concatenated. For example, the embedding of an image would be generated by an AI model trained on images, the embedding of the text describing the image would be generated by an AI model trained on texts, and then the two embeddings would be concatenated to represent the image and its transcript. Instead, the Gemini family of models was trained on content that is inherently multimodal such as text, images, videos, code, and audio. Imagine being able to ask a question about a picture, or generate a poem inspired by a song – that’s the power of Gemini.
** #Mixture-of-Experts Model: At the core of Gemini’s groundbreaking capabilities lies its innovative mixture-of-experts model architecture. Unlike traditional neural networks that route all inputs through a uniform set of parameters, the mixture-of-experts model consists of numerous specialized sub-networks, each adept at handling different types of information or tasks—these are the “experts.” Upon receiving an input, a gating mechanism intelligently directs the input to the most relevant experts. This selective routing allows the model to leverage specific expertise for different aspects of the input, akin to consulting specialized departments within a larger organization for their unique insights. For Gemini, this means an unparalleled ability to process and integrate a vast array of multimodal data—whether it’s textual, visual, auditory, or code-based—by dynamically engaging the most suitable experts for each modality. The result is a model that not only excels in its depth and breadth of understanding but also in computational efficiency, as it can focus its processing power where it matters most, without overburdening the system with irrelevant data processing. This approach revolutionizes how AI models handle complex, multimodal inputs, enabling more nuanced interpretations and creative outputs than ever before.
*** #Reasoning: Gemini goes beyond simple pattern recognition. It utilizes a novel architecture called “uncertainty-routed chain-of-thought” to reason and understand complex relationships within and across modalities. This enables it to answer open-ended questions, solve problems, and generate creative outputs that are not just factually accurate but also logically coherent.
🤖 #GenAI everyday hacks: 📊 #Comparison: How to create a handy cheatsheet for the 2024 Oscars?
🎬👩🎤 Excited for the 96th Academy Awards this Sunday, March 10th? Are you like me and wondering: ❓Which movie scored the highest number of nominations? ❓How many Best Picture contenders scored both Leading Actress and Leading Actor nominations? ❓Is there a movie nominated in all key categories (Best Picture, Directing, Leading Actress and Actor, Writing)
Let’s turn this data into insights with #GenAI! Acting as a #PromptEngineer, I ran an experiment with a few AI tools, such as #Gemini, #chatgpt, #copilot and #perplexity etc. Here’s the outcome: ✅ GenAI can help! 🙀 But it may take several prompts to get the results you want.
I started with this prompt: “Compare the Best Picture nominees for the 2024 Oscars in a table ordered by the total number of nominations received and specifying if the movies were nominated in the Directing, Leading Actor, Leading Actress and Writing categories.”
After a couple of additional prompts, including: “Specify if Writing as Adapted or Original Screenplay.” and “Replace ‘Yes’ with the name of the nominee and ‘No’ with ‘_'”, and a few “Try again”, I got the table below.
✅ In a few minutes I got a great 🎬 2024 Oscars cheatsheet. 🙀 But it took more prompts than I expected. For example, despite my clear request for a table, several of my prompts returned a text based response, so I had to ask my AI bots to “Try again”.
Lessons learned: 🕵♀️ Refining your prompts sequentially enhances AI’s ability to deliver complex content, so plan for some trial and error. 🕵♀️ AI bots don’t always want to do the heavy lifting on the first ask so be persistent. 🕵♀️ Always check the results to ensure #EmmmaStone does not become #MargotRobbie
The AI tools are the talented but sometimes stubborn #cast👩🎤 and the #PromptEngineer is the director 🎥
🤖 #GenAI everyday hacks: 📕 #Summarization: How to turn a daunting 150+ page-long voter pamphlet into a handy cheatsheet for the SF Elections?
✅ GenAI can help! 🙀 But your mileage will vary. Let’s see an example.
I used this prompt: “Summarize all the propositions on the March 5th, 2024 San Francisco Election ballot with their top arguments for and against in a 3-column table using the voter pamphlet as the data source.”
After a couple of additional prompts, such as: “You forgot prop A.” and “Make it more concise.” I got the table below.
✅ In a couple of minutes I got a great SF Elections cheatsheet. 🙀 But some fields are more accurate then others. For example, there are no arguments against Proposition G in my voter pamphlet, and so “Curriculum pressure on students and teachers.” field is entirely made up.
Lesson learned: 🕵♀️🕵♂️ Always verify your AI-generated content.
Amazingly, I can fit my key program learnings into one sentence, which is a testament to how incredibly well Baba Shiv (one of my absolute favorite Stanford professors!) designed it alongside the program director Angel Dodson:
Acting and engineering processes have much in common: ✅ Requirements Gathering 👷♀️🛠️ is like Character Development 👨🎤🎬 ✅ Development 👷♀️🛠️ is like Preparation 👨🎤🎬 ✅ Deployment 👷♀️🛠️ is like Showtime 👨🎤🎬
Both crafts require essential tools that together drive technology innovation:
🔵 #Precision (Solve for the right “it”) defines vital innovation scope and focus. But without #Clarity (What’s your role? Who’s your audience?) even the greatest innovation may not deliver user value.
🔴 #Prototyping (Build ▶ Assess ▶ Improve; Repeat) allows you to pivot quickly and accelerate progress. But without #Storytelling (connecting emotionally with your audience) you risk stakeholder buy-in and your innovation will never scale.
⚫ #Premortem (Preemptively address risks) will help you avoid both the common pitfalls and less common risks. But without #Presence (Your best performance, adjusted on the spot) you may fail to read your audience and overlook subtle cues in feedback from your board, investors, or customers ahead of the launch.
So there you have it: the six tools in the Innovative Technology Leader’s toolbox. Is any of the six tools surprising to you?
Starting on the way to our morning exercise classes (5:45am! 😱 🏃♀️ 🏃♂️ ) and all the way into our evening receptions ( 🍷 🥗 ), AI monopolized 90% of our conversations. And for good reasons. You can benefit from Generative AI in multiple ways today.
On the 4th day of the program, I white-boarded a prototype of the gen AI #ValueSpectrum to visualize the range of use cases where gen AI adds value today, and how you can unlock it. In this post, I’m sharing a more polished version.
Who is the CEO that started their company in a California garage in 1945, was an avid risk taker, a glutton for data, obsessed with releasing new products every year and aggressive in adopting technology? Nope, it’s not a Silicon Valley tech pioneer, but the incredible Ruth Handler, the founder of Mattel, Inc. Talk about thinking outside the toy box! 😀
This past weekend, The Wall Street Journal and The New York Times celebrated the release of #barbiethemovie with stories on the legendary creator of the iconic toy, and I must say, Ruth Handler is my new marketing idol, innovation inspiration, and feminist icon all rolled into one! 🙌
#Marketer: Ruth turned the toy market upside down. She was a TV advertising trailblazer when others were still figuring out newspaper ads! 📺 Forget selling toys just before the holidays, Ruth said, “Let’s make every day a play day!” 📅 And who says parents make all the decisions? Not Ruth! She marketed directly to kids, and it was genius. 🎯 https://lnkd.in/e4eNkaCi
#Innovator 💡 Talk about a lightbulb moment! Ruth came up with the Barbie idea while her daughter, Barbara, was playing with paper dolls. Little Barbie Handler and her friend went on a wild imagination spree, dressing the cutouts in different outfits, dreaming up careers, and personalities! Sounds like a Stanford University Graduate School of Business case study on innovation to me! 📚 https://lnkd.in/eg73AnJ4
#Feminist 💪 Barbie wasn’t just another baby-doll encouraging girls to play mom; she was the ultimate role model! She went to the moon before Neil Armstrong even got close, and she’s been president – something no real-life American woman has done yet! Breaking barriers like a boss! 👩🚀 https://lnkd.in/eBmEZy2Y
I’m raising my tiny Barbie-sized coffee cup to Ruth Handler – the ultimate maven of marketing, the queen of innovation, and a true feminist trailblazer! 👑
See the full post on LinkedIn and share your thought via comments here.
I’m a Marketing Executive, an astute influencer, panelist, and public speaker with recent appearances on Harvard Business Review (HBR). I live and work in San Francisco.
4Simone Morellato and 3 others
Tech Surprises 