A recent MIT report revealing that 78% of businesses face challenges in AI adoption due to weak data foundations. Despite high aspirations for AI, such as enhancing efficiency and fostering innovation, only a small percentage of business leaders feel adequately prepared to engage with AI technologies. The report highlights that data governance and quality are significant obstacles, emphasizing the need for organizations to address underlying data issues before embarking on AI projects.
The episode also covers the FIDO Alliance's announcement at the Authenticate conference regarding the Credential Exchange Protocol (CXP), which aims to make passkeys portable across different digital ecosystems. This initiative, supported by major tech companies like Apple, Google, and Microsoft, seeks to standardize secure transfer processes for passkeys, reducing concerns over vendor lock-in. The growing acceptance of passwordless technology is underscored by Amazon's report of over 175 million customers activating passkeys, indicating a shift towards more secure digital practices.
Host Dave Sobel contrasts the approaches of OpenAI and Anthropic in the realm of AI development. OpenAI's new O1 model family focuses on structured prompt generation for efficiency and accuracy, while Anthropic emphasizes a more conversational and engaging AI experience. This divergence in methodologies reflects the differing priorities of the two companies, with OpenAI aiming for structured outputs and Anthropic fostering a personable interaction style. Additionally, a study by Apple researchers reveals limitations in mathematical reasoning within large language models, stressing the importance of human oversight in decision-making processes.
Finally, the episode highlights the need for organizations to modernize their cybersecurity strategies by adopting an identity-centric zero trust model. As hybrid work environments and AI technologies challenge traditional security practices, a zero-trust approach ensures that all access requests are verified, protecting sensitive resources from unauthorized access. Sobel also discusses the importance of fostering AI experimentation within organizations to bridge the gap between individual productivity gains and overall performance, encouraging a cultural shift towards embracing AI as an integral part of business strategy.
Four things to know today
00:00 78% of Businesses Struggle with AI Adoption Due to Weak Data Foundations, Says MIT Report
02:12 FIDO Alliance Unveils New Protocol to Make Passkeys Portable Across Platforms
03:56 OpenAI’s Structured AI vs. Anthropic’s Conversational AI: What It Means for Business Efficiency
06:23 Why AI Experimentation and Zero-Trust Security Are the Future of Organizational Strategy
Supported by: https://mspradio.com/engage/
Event: www.smbTechFest.com/Go/Sobel
💼 All Our Sponsors
Support the vendors who support the show:
👉 https://businessof.tech/sponsors/
🚀 Join Business of Tech Plus
Get exclusive access to investigative reports, vendor analysis, leadership briefings, and more.
👉 https://businessof.tech/plus
🎧 Subscribe to the Business of Tech
Want the show on your favorite podcast app or prefer the written versions of each story?
📲 https://www.businessof.tech/subscribe
📰 Story Links & Sources
Looking for the links from today’s stories?
Every episode script — with full source links — is posted at:
🎙 Want to Be a Guest?
Pitch your story or appear on Business of Tech: Daily 10-Minute IT Services Insights:
💬 https://www.podmatch.com/hostdetailpreview/businessoftech
🔗 Follow Business of Tech
LinkedIn: https://www.linkedin.com/company/28908079
YouTube: https://youtube.com/mspradio
Bluesky: https://bsky.app/profile/businessof.tech
Instagram: https://www.instagram.com/mspradio
TikTok: https://www.tiktok.com/@businessoftech
Facebook: https://www.facebook.com/mspradionews
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
[00:00:02] It's Thursday, October 17th, 2024, and I'm Dave Silver. Four things to know today.
[00:00:07] 78% of businesses struggle with AI adoption due to weak data foundations, says an MIT report.
[00:00:13] The FIDO Alliance unveils a new protocol to make Passkeys portable across platforms.
[00:00:19] OpenAI's Structured AI versus Anthropics' conversational one, what it means for business efficiency, and
[00:00:25] why AI experimentation and Zero Trust Security are the future of organizational strategy.
[00:00:32] This is the Business of Tech.
[00:00:35] A recent report by MIT Technology Review Insights and Snowflake reveals that 78% of organizations
[00:00:41] struggle to leverage AI effectively due to inadequate data foundations.
[00:00:46] Despite high aspirations for generative AI, 72% aim to enhance efficiency, 55% seek increased
[00:00:53] competitiveness, and 47% hope for innovation, only 22% of business leaders feel very ready
[00:01:00] to engage with AI. The report indicates that 95% of those surveyed face challenges in AI implementation,
[00:01:08] with data governance and quality being the most significant obstacles.
[00:01:12] Barris Gutlin, head of AI at Snowflake, emphasizes that a robust data foundation is essential for unlocking
[00:01:19] AI's potential, enabling businesses to transform operations and products while addressing security
[00:01:24] and cost concerns. Why do we care? The fact that 78% of organizations struggle to leverage AI effectively
[00:01:32] due to weak data foundations underscores a fundamental truth. AI is only as good as the data it is built on.
[00:01:39] While companies have high hopes for what AI can achieve, many lack the necessary infrastructure and governance
[00:01:45] to support these ambitions. The core problem here isn't necessarily with AI itself. It's that most businesses aren't ready
[00:01:52] to operationalize AI because they haven't solved their underlying data issues. This is a crucial insight
[00:01:58] for IT service providers, consultants, and business leaders alike. Before embarking on AI projects,
[00:02:04] organizations need to address data silos, governance, and integration to create a scalable foundation.
[00:02:13] At the FIDO Alliance's Authenticate conference, researchers announced two significant projects to advance
[00:02:19] passkeys as a secure alternative to traditional passwords. The first is the Credential Exchange Protocol
[00:02:25] or CXP, designed to enable the portability of passkeys across different digital ecosystems, addressing concerns
[00:02:32] over user lock-in. This initiative is backed by major players in the tech industry, including Apple, Google,
[00:02:38] and Microsoft, and aims to standardize secure transfer processes for passkeys and other sensitive data.
[00:02:45] A new resource website, Passkeys Central, will provide developers with essential tools and documentation
[00:02:50] to implement passkeys effectively. According to Andrew Shakar, CEO of the FIDO Alliance,
[00:02:56] the collaboration among competitors signifies a unified effort to reduce dependency on passwords
[00:03:02] with growing adoption of passkeys across various platforms.
[00:03:05] Craig Newmark Philanthropies also supports these advancements, emphasizing the need for improved
[00:03:10] digital security against increasing threats.
[00:03:13] Major companies, including 1Password and Dashlane, have already committed to supporting these new standards,
[00:03:19] which are open for public review before final adoption.
[00:03:22] Additionally, Amazon reported that over 175 million customers have activated passkeys for their accounts,
[00:03:28] highlighting the growing acceptance of this passwordless technology.
[00:03:33] Why do we care?
[00:03:34] The move away from passwords has long been necessary, but one of the hurdles has been fear of vendor lock-in.
[00:03:40] With CXP, IT service providers and businesses can confidently invest in passkey infrastructure,
[00:03:46] knowing that the risk of being tied to a single vendor is reduced.
[00:03:49] This paves the way for more scalable, passwordless solutions.
[00:03:53] And that's why we care.
[00:03:56] This pair of stories was almost a big idea.
[00:03:59] OpenAI has unveiled its meta prompt for the new O1 model family,
[00:04:04] focusing on structured prompt generation to enhance user actions with its products.
[00:04:09] Unlike Anthropix Claude, which emphasizes a more conversational and human-like approach,
[00:04:14] OpenAI's system is designed for efficiency and accuracy, treating its AI as a powerful computational tool.
[00:04:20] The meta prompt includes detailed instructions for logical reasoning and problem solving,
[00:04:26] encouraging a step-by-step breakdown of tasks, a feature that enhances its usability for complex problems.
[00:04:32] In contrast, Anthropix's approach fosters a more friendly and engaging AI,
[00:04:37] allowing for a narrative style that reflects personality traits.
[00:04:41] This divergence in methodologies highlights the differing priorities of the two companies.
[00:04:46] OpenAI prioritizes structured outputs for task completion.
[00:04:49] Anthropix focuses on creating a more interactive and personable AI experience.
[00:04:55] And a recent study by Apple researchers, led by Sammy Bengio and Onsel Tuzel,
[00:05:00] reveals significant limitations regarding mathematical reasoning in large language models.
[00:05:06] Their paper, GSM's Symbolic, Understanding the Limitations of Mathematical Reasoning in Large Language Models,
[00:05:12] introduces a new dataset for evaluating LLMs.
[00:05:15] It suggests that typical benchmark datasets, like GSM 8K, may present an overly optimistic view of AI capabilities.
[00:05:25] The study found that as questioned complexity increases, accuracy can drop dramatically from the high 80-90% range to as low as 40%.
[00:05:34] This emphasizes the importance of human oversight in decision-making processes, as LLMs excel in pattern matching but struggle with logical reasoning.
[00:05:44] The researchers urge businesses to cautiously invest in AI technologies, highlighting that while AI can enhance operations, it should not replace human expertise.
[00:05:54] Why do we care?
[00:05:56] As the landscape of AI continues to evolve, understanding these differences may help users navigate their interactions with these models more effectively.
[00:06:04] Businesses must be savvy in how they adopt and apply AI.
[00:06:08] This involves selecting the right tools and ensuring they are deployed where they can deliver real value, all while maintaining human oversight when necessary.
[00:06:17] Critical thinking, logical reasoning, and decision-making tasks still require human expertise.
[00:06:24] It's time for some big ideas.
[00:06:27] As I prepare for some upcoming interviews, I've been discussing identity extensively.
[00:06:32] So this one from Dark Reading.
[00:06:34] Why your identity is the key to modernizing cybersecurity.
[00:06:37] Organizations must modernize their cybersecurity strategies in a rapidly evolving digital landscape by adopting an identity-centric zero-trust model.
[00:06:46] This shift is crucial as hybrid work environments and AI technologies challenge traditional security practices, which rely on a moat and castle approach.
[00:06:54] As employees increasingly work remotely and access cloud services, a zero-trust model mandates that all access requests are verified, ensuring that only authorized users and devices can connect to sensitive resources.
[00:07:07] Without such measures, deploying advanced AI tools poses significant risks, as unauthorized users could easily exploit these systems to access confidential information.
[00:07:17] An e-book I'll make you aware of, The S&B Technology Buying Journey, Part 2, by the S&B Group.
[00:07:23] Based on recent survey data, the e-book highlights critical aspects of how S&B's research evaluate and select technology solutions.
[00:07:32] Key findings include preferred information sources, content preferences, and the extent of self-guided research completed before engaging vendors.
[00:07:41] Additionally, it examines the top purchase channels for software and infrastructure technologies, as well as the preferred methods of communication from vendors post-sale.
[00:07:51] The resource provides technology vendors with insights to target S&B customers better and guides S&Bs to navigate the complex technology purchasing process.
[00:08:00] The e-book is available for a free download.
[00:08:03] If you want to make AI more practical, how about some tactics?
[00:08:06] In a recent article by Ethan Mollick, it was highlighted that a significant number of workers are utilizing AI in their jobs, with a study from Denmark showing that 65% of marketers and 64% of journalists have engaged with AI tools.
[00:08:20] A separate study in the U.S. revealed that one-third of workers used generative AI within a week, with productivity improvements of 25% noted in consulting tasks and a 26% increase from the original GitHub co-pilot for coding.
[00:08:34] Despite these advancements, many organizational leaders report minimal AI use, indicating a disconnect between individual productivity gains and overall organizational performance.
[00:08:45] Mollick suggests companies must focus on internal research and development initiatives to bridge the gap, encouraging experimentation among employees while alleviating fears of repercussions.
[00:08:56] Why do we care?
[00:08:57] There's some great strategy here, fostering AI experimentation and cultural buy-in at an organizational level, rather than treating AI as a bolt-on tool.
[00:09:06] Aligning sales and marketing strategies to match the evolving tech purchasing journeys of small businesses, focusing on content and relationship building.
[00:09:14] And zero trust models for both security and AI deployment.
[00:09:17] Some big ideas for you to consider.
[00:09:21] Looking to reach an audience of thousands of MSPs and IT service providers?
[00:09:25] Put your ad right here on the Business of Tech, and be on the show that 64% of MSPs report having listened to.
[00:09:34] A recurring Top 50 tech news podcast, there are affordable options for you to reach our audience, and we can support any budget.
[00:09:42] Podcast listeners are more engaged, have a higher level of brand retention, and are more willing to listen to ads here than any other avenues.
[00:09:53] Want to know more?
[00:09:54] There's information at mspradio.com slash engage, including a button to book a time to talk.
[00:10:01] I'm looking forward to that discussion.
[00:10:05] Thanks for listening.
[00:10:06] It's National Clean Up Your Virtual Desktop Day, so get that stuff off your background.
[00:10:10] It's also National Pasta Day.
[00:10:13] Tomorrow, you'll get a bonus episode as I talk with Stephen Cook, owner of Strategic IT Services.
[00:10:18] I enjoy talking to MSPs to get insights from the field.
[00:10:22] Stephen says he has a hot take.
[00:10:23] I'll leave it to you to judge his thoughts.
[00:10:26] Then on Sunday, you get my interview with Avi Perez, the CEO and co-founder of Pyramid Analytics.
[00:10:33] Some great insights into data governance.
[00:10:36] I'll be back again on Monday.
[00:10:38] If you got a comment or a thought on a story, put it on the comments if you're on YouTube or reach out on LinkedIn if you're listening to the podcast.
[00:10:43] Have a great weekend.
[00:10:45] The Business of Tech is written and produced by me, Dave Sobel, under ethics guidelines posted at businessof.tech.
[00:10:52] If you like the content, please make sure to hit that like button, follow or subscribe.
[00:10:58] It's free and easy and the best way to support the show and help us grow.
[00:11:03] You can also check out our Patreon where you can join the Business of Tech community at patreon.com slash MSP radio or buy our Why Do We Care merch at businessof.tech.
[00:11:16] Finally, if you're interested in advertising on this show, visit MSP radio dot com slash engage.
[00:11:22] Once again, thanks for listening to me.
[00:11:25] I'll talk to you again on our next episode of the Business of Tech.
[00:11:32] Part of the MSP radio network.
[00:11:35] .
[00:11:35] .

