New Analysis: Nvidia GTC 2024 Is The Davos of AI

R "Ray" Wang

Principal Analyst and Founder
Constellation Research

R “Ray” Wang is the CEO of Silicon Valley-based Constellation Research Inc. He co-hosts DisrupTV, a weekly enterprise tech and leadership webcast that averages 50,000 views per episode and blogs at www.raywang.org. His ground-breaking best-selling book on digital transformation, Disrupting Digital Business, was published by Harvard Business Review Press in 2015. Ray's new book about Digital Giants and the future of business, titled, Everybody Wants to Rule The World was released in July 2021. Wang is well-quoted and frequently interviewed by media outlets such as the Wall Street Journal, Fox Business, CNBC, Yahoo Finance, Cheddar, and Bloomberg. Short Bio R “Ray” Wang (pronounced WAHNG) is the Founder, Chairman, and Principal Analyst of Silicon Valley-based Constellation Research Inc. He…...

Read more

New Chips, New Business Models Ahead In Accelerated Computing

Nvidia stock is up 267% YoY.  With over 10,000 people in physical attendance and up to 100,000 people virtually attending, this could be the largest AI conference. All eyes on the next wave for AI with CEO Jensen Huang ready to keynote at 1:00 pm PT at the SAP Center in San Jose.  The key things to look for in the keynote:

• New B100 Chip replacement.  Built by TSMC, the new chips are expected to be advanced 3nm design.  NVIDIA will take advantage of the two dieTSMC's CoWoS-L (Chip-on-Wafer-On-Substrate-L) advanced 2.5D packaging technology to get larger processors.  Should Nvidia move from the monolithic design to a multi-chip module (MCM) approach, the chip maker could create faster derivatives and shorten its product cycle.
• CUDA advancements. This is the software layer that allows software companies to harness the power of the GPUs
• Key vertical expansion.  Healthcare, pharma, defense, and public sector consume massive workloads.
• Partnerships across the AI ecosystem. Expect announcements from partners such as OpenAI, Microsoft, Amazon, and Google, Meta, Micron, Oracle, Super Micro, Dell, Intel, and more.
• Sovereign AI growth.  Nvidia is focusing beyond data centers and serving governments for sovereign AI.

Chips Are The Foundational First Inning Of A Nine Inning Age Of AI

The winners of the internet weren’t the early infrastructure companies They were the companies that built their business model on the internet as a key distribution channel.  In AI, it will be the companies that take advantage of the full-stack of AI.  Software companies will drive the second inning.  But the third and fourth innings will be the innovative companies who use AI as their core business model.  For example, in the war for refrigeration, it wasn't the appliance manufacturers that won, it was actually Coca Cola that innovated a new market for cold beverages.

What if a Trillion PDFs Could Talk?

Liz Miller

Vice President & Principal Analyst
Constellation Research

About Liz Miller: Liz Miller is Vice President and Principal Analyst at Constellation, focused on the org-wide team sport known as customer experience. While covering CX as an enterprise strategy, Miller spends time zeroing in on the functional demands of Marketing and Service and the evolving role of the Chief Marketing Officer, the rise of the Chief Experience Officer, the evolution of customer engagement, and the rising requirement for a new security posture that accounts for the threat to brand trust in this age of AI. With over 30 years of marketing experience, Miller offers strategic guidance on the leadership, business transformation, and technology requirements to deliver on today’s CX strategies. She has worked with global marketing organizations to transform everything from…...

Read more

No matter where you turn these days, you are bound to hear about a little something called generative AI. I’ve said it before…AI is really a “thing” now. Don’t misunderstand my sarcasm: Generative pretrained transformers are not new. Neither is artificial intelligence or the models that power them. But within the last year, we’ve seen leaps and bounds in how these innovations are being brought together in the name of productivity.

Case in point: The AI Assistant in Adobe Acrobat and Reader. That’s right…your PDF just got chattier.

What’s the News Here?

Adobe has launched a beta of an AI Assistant that brings conversational and generative AI capabilities to Reader and Adobe Acrobat. Essentially, this new “conversational engine” can summarize the content of a document supported by Acrobat or Reader, answer questions, and recommend more questions based on the content. The models behind Acrobat Liquid Mode bring responsive reading for PDFs on mobile devices, and Adobe’s unique understanding of different document file formats and structures boosts the quality of responses, output and recommendations.

In researching this post, I have played with the beta, and I have to agree that the responses are accurate, but also noted that often the answers were supplemented with citations and additional summaries and insights from across a couple very lengthy documents. I found that reading a document and also leveraging the AI Assistant got me to dig deeper into some thoughts which was something I wasn’t necessarily expecting from the experience.

In the beta, Adobe Acrobat and Reader now include:

• AI Assistant: Answers questions directly from users in a highly interactive and conversational manner, recommends and delivered suggested and custom questions based on the content of the document.
• Generative summary: short overviews of content in easy-to-read format.
• Intelligent citations: attribution engine and generative AI created citations to verify the source of AI Assistant’s summaries and answers to user questions.

The AI Assistant beta is currently available, allowing Acrobat Reader and Acrobat Individual, Pro and Teams customers to simply open their applications and start chatting with their documents today. There is also a private beta for enterprise customers. New features will be coming over the next several weeks. For the time being, these features are available on desktop and web. English is currently supported, but Adobe notes that other languages are soon to follow. Once out of beta, Adobe plans to release a new add-on subscription plan for the full range of AI Assistant capabilities. Currently, there is no word on how long AI Assistant is expected to be in beta, but there is already an extensive roadmap of future capabilities.

What’s the Impact Here?

Adobe’s AI Assistant brings the summary and productivity power of AI to the gold standard in business documentation. I struggle to think of the last time a white paper, contract or user manual wasn’t delivered in a large, multi-page PDF. Now…those PDFs can talk. Or rather…they can listen and then talk back with these new conversational AI functions. The impact on productivity and usability of documents is unquestionable. Imagine being able to just ask product guide how to troubleshoot, or pinpoint exactly where specific phrases or critical details like pricing are in a document. Imagine being able to ask if there are any differences in that pricing. Teams can start to ask for summaries of trends, send meeting notes or chapter summaries, or ask AI to identify anomalies in documents like contracts.
 

What Makes Me Lean in Here?

While the usability of the AI Assistant is unquestionable, we must think about how we as content marketers and creators need to plan and perhaps even write differently thanks to AI Assistants and copilots like this one from Adobe. As a content marketer, things like chapters, bookmarks, clickable links, sticky notes, and comments are lifeline tools when using a PDF as the delivery mechanism for your thought leadership. Afterall, you don’t just “export as PDF” and call it a day.

In this age of Generative AI, we will need to think differently and create within new mental models. When I launched Adobe Acrobat and opened a recent report I had written, I was greeted by the AI Assistant who “knew me” by name and it said, “I’m here to help you read, understand and work with your documents more efficiently.” I must admit that I was left wondering if I had set up the right phrasing, keywords, summary statements and guides to help the AI Assistant summarize and understand the big takeaways that I intend readers to walk away with. It made me exceedingly aware that I needed to not just think about the thoughts I was conveying but HOW I was structuring those thoughts to work well with generative AI tools like this assistant…was I writing to be seen and summarized?

Generative AI will open a world of content possibilities, even offering to author emails, change tone and tenor, update information based on pricing, product or even brand details. It will also offer to break through my writer’s block and whip up that first draft for me…even offer to edit or tighten up my run on sentence. So how can I be sure that I’m clear in those thoughts or leaving the right breadcrumbs to new thoughts?

Thought leadership content is hard enough to create as it needs to be something truly new with thinking that stands the test of time. Now we must make sure that these new thoughts are understood and summarized as such. We won’t be able to hide from the spotlight generative AI can shine on old ideas masquerading as new big ideas!

Content strategy and an overall understanding of where and how AI is impacting an organization’s capacity to create, contextualize and individualize content has to modernize and change. Understanding how these new AI powered tools are fundamentally shifting the process and operations of content marketing has never been MORE important.

With Adobe’s new AI Assistant, PDFs could be saying a lot more than we ever intended them to say. Are we as content marketers ready for this new conversation?

For more on the new AI Assistant, check out Adobe’s announcement here: https://news.adobe.com/news/news-details/2024/Adobe-Brings-Conversational-AI-to-Trillions-of-PDFs-with-the-New-AI-Assistant-in-Reader-and-Acrobat/default.aspx

Header Image was AI Generated using Adobe Firefly

Salesforce Reinvents Itself Again With Einstein 1

Chirag Mehta

Vice President and Principal Analyst
Constellation Research

Chirag Mehta is Vice President and Principal Analyst focusing on cybersecurity, next-gen application development, and product-led growth. With over 25 years of experience, he has built, shipped, marketed, and sold successful enterprise SaaS products and solutions across startups, mid-size, and large companies. As a product leader overseeing engineering, product management, and design, he has consistently driven revenue growth and product innovation. He also held key leadership roles in product marketing, corporate strategy, ecosystem partnerships, and business development, leveraging his expertise to make a significant impact on various aspects of product success. His holistic research approach on cybersecurity is grounded in the reality that as sophisticated AI-led attacks become…...

Read more

It’s Albert Einstein’s birthday today. Einstein would not have imagined 885 people dressing up as him to set a Guinness World Record, 145 years after he was born. Salesforce held a special event last week in San Francisco, during their developer-focused conference, TrailblazerDX, where Trailblazers dressed up as Einstein to break a Guinness World Record. During this conference, Salesforce announced their next-generation AI platform, Einstein 1 Studio. The energy on the show floor was palpable as hundreds of developers hurried from expert sessions to demo stations to the Hacker Camp. Thousands of other developers tuned in online to stream the conference programming.

Over the course of years, Salesforce has made many bold decisions to continue to reinvent itself. On its 25^th anniversary this week, Salesforce once again reinvented itself by deciding to go all in with AI. When Parker Harris, co-founder of Salesforce, took the keynote stage to announce Einstein 1, he joked about him looking old. He was as energetic and beaming with enthusiasm as one can be. On a beach-themed stage, he even threw a beach ball to the next presenter after flawlessly pronouncing his name in single breath, Muralidhar Krishnaprasad, the engineering head of Data Cloud, who goes by MK. Parker also took the opportunity to remind the audience about Salesforce’s 1-1-1 model: donating 1% equity to charitable causes, 1% products to non-profits, and 1% of employees’ time in community service.

Einstein 1 Studio includes Prompt Builder, Copilot Builder, and Model Builder—three essential ingredients of a next-generation platform. Salesforce’s post captures details on these specific components. At high-level, the platform caters to developers who want to build AI-driven experiences that can leverage Salesforce metadata, customizable workflows, and underlying predictive and generative AI-models that are either trained inside Salesforce or elsewhere. These Salesforce developers also want to maximize code reuse and program a system using a natural language. The foundation of Salesforce’s big bet on AI is its fast-growing Data Cloud. MK talked up the role of Data Cloud in preventing hallucination and keeping Einstein 1 “grounded in trust.”

My insights

With Einstein 1, Salesforce is in a unique position to serve customers in three important dimensions—use data to strengthen AI, provide a platform to build and extend product experiences, and provide first-party and third-party solutions. Salesforce has clearly demonstrated their intent to lead their platform with AI. They are also matching their intent with investment in various areas, products as well as the underlying platform.

As Salesforce blazes forward with Einstein 1, and as CxOs closely navigate this fast-moving chaotic domain to make informed decisions about their own investment in AI, here are a few things worth highlighting:

Ecosystem beyond Salesforce

Einstein 1 platform’s near-term roadmap is designed to help existing Salesforce customers to get the best out of their investment. There are many more prospects, however, who don’t use Salesforce today, their products or the platform. These prospects are exploring their next-gen platform choices that include platforms from hyperscalers, niche AI vendors, and select global service integrators (GSI). Many of these prospects have also invested in various copilots. While Einstein 1 can clearly compete to win, it’s unclear whether it is enough of a draw for these prospects to consider Salesforce. As copilots proliferate, customers want to avoid an interoperability nightmare. While most vendors do offer lower-level API integrations, lack of well-defined semantics and interfaces pose interoperability challenges for components that operate at higher levels.

And then there are large number of startups (ISVs) that have choices to make. Clara Shih, CEO of Salesforce AI, underscored the value proposition of Einstein 1 on why startups should consider building on Salesforce instead of investing their limited resources in building core AI components that Salesforce is heavily investing in. ISVs will continue to be the one of the biggest customers of any next-gen platform. If ISVs can get what they need from Salesforce, and if Salesforce can provide them effective distribution and commerce via AppExchange, there is an opportunity to enable a brand-new AI ecosystem.

Benchmarking next-gen platforms

While engineering and business efficiency gains from using a next-gen platform such as Einstein 1 are very real, the industry lacks KPI definitions and a benchmark to compare these platforms and measure ROI. It’s unclear what makes a copilot a great copilot or what makes a prompt designer a great prompt designer. Enterprise software vendors do use domain-specific KPIs such as time to close a ticket, increase in pipeline coverage etc., but IT leaders want to compare various platforms to better understand their ROI. While this is clearly not vaporware, IT leaders have been burnt before by making an early investment in technology they did not fully understand. There will always be a qualitative comparison, but a good set of KPIs ground everyone in facts. It’s time someone takes a lead to define what these KPIs could look like.

Bridging PX and DX

A promise of a next-gen platform—the one that goes beyond code reusability, productivity gains, and natural language driven development—is how fast one can go from an idea or a sketch to a working application with customer validation. AI can effectively help bridge the gap between product experience (PX) and developer experience (DX) with faster feedback loops. Finding the right problem to solve is equally, if not more, important than solving it efficiently. The last thing you want to use AI for is to increase the efficiency of solving a wrong problem. Salesforce’s Flow Builder is the closest to a visual representation of a working application but it’s not a prototyping tool. Salesforce and the rest of the industry have a unique opportunity to increase their focus upstream—investing in AI-driven declarative prototyping tools—to reap the benefits downstream, faster time to market and time to value.

Happy coding, or shall I say happy generating!