Welcome to The Daily AI Show Newsletter, your deeper dive into AI that goes beyond the latest news. In this issue:

• Can AI Really Discover, or Just Repackage Our Knowledge?

• Beyond School: How AI Tutors Could Reshape Careers and Personal Growth

• Taking the Tension Out of Big Decisions—With a Little AI Help

• Can ChatGPT’s New Search Feature Outshine Perplexity?

Plus, we discuss Amazon Prime’s new X-ray Recap, if Sam just said AGI will happen in 2025, Microsoft’s not-so-subtle plot to get home 365 users to upgrade to CoPilot, Advanced Search in ChatGPT Desktop, AI’s good and bad impact in the US election, if AI being fully representative means homogenization, and all the new stories that caught our eyes and ears this past week.

It’s Sunday morning.

Black Friday is coming, and the algorithms have already decided which deals you’ll click on.

But hey, you can still use your remaining free will to read this week’s edition.

. . . .or did our AI already know you would. 🤯

The DAS Crew

Can AI Really Discover, or Just Repackage Our Knowledge?

Inspired by Dan Shipper’s Chain of Thought article, this thought experiment questions whether AI can genuinely create new knowledge—or if it’s limited to rearranging what humans already know. Imagine an AI model trained solely on texts from the 1500s. Would it ever arrive at principles like gravity or modern physics on its own, or would it remain confined by the boundaries of historical understanding?

Human curiosity drives us to explore the unknown and find unexpected connections. While AI excels at spotting patterns and reassembling ideas in new ways, true discovery may require more than computation. Shipper’s article challenges us to consider the limits of machine learning:

Will AI ever cross the line from mimicry to original thought?

WHY IT MATTERS

Understanding AI’s Limits in Your Business: Knowing that AI’s creativity is rooted in remixing existing knowledge helps set realistic expectations. It’s a powerful tool for productivity and insight, but true leaps in strategy or product innovation might still rely on human creativity.

Accelerating—Not Replacing—Human Insight: AI can help surface patterns and suggest optimizations, making it an invaluable asset for research and efficiency. However, its role is to augment your expertise, not replace it.

AI as a Discovery Partner, Not a Replacement: When implementing AI, focus on using it as a tool to accelerate discovery within existing knowledge frameworks. It’s excellent for spotting trends and refining ideas but might need your input to break truly new ground.

Potential for Problem Solving in Your Field: AI’s strength in synthesis means it can offer novel solutions by recombining known elements—ideal for problem-solving in areas like customer support, operations, and data analysis.

Guiding AI with Clear Goals and Input: Since AI’s “knowledge” is derivative, guide its application with clear objectives and domain expertise. This way, you’re using AI’s potential effectively without expecting it to “think” independently.

Beyond School: How AI Tutors Could Reshape Careers and Personal Growth

AI-based tutoring is proving to be a game-changer not only in education but also in professional and personal development. A recent small study at Harvard demonstrated that students using AI tutors learned twice as much in half the time, highlighting the potential of AI to offer customized learning experiences.

This isn't limited to school; AI tutors and assistants are extending their benefits to workplaces, supporting employees in skills training, knowledge retention, and continuous professional development.

As AI becomes a more integral part of daily life, personalized tutors can help individuals at all stages—students, professionals, and lifelong learners—gain mastery over subjects or skills in a structured, self-paced format.

For businesses, AI-powered training programs offer a scalable way to upskill employees, keeping them agile in a rapidly evolving marketplace. The accessibility of such tools also means that industries, particularly those with high turnover or specific knowledge requirements, can create a more knowledgeable, adaptable workforce without the need for costly, time-intensive training programs.

WHY IT MATTERS

Accelerated Skill Development: AI tutors may speed up learning, making it easier for professionals to quickly build new skills or deepen expertise, which is essential for staying competitive in fast-changing industries.

Improved Knowledge Retention in the Workplace: By enabling individualized learning experiences, AI enhances knowledge retention, a critical advantage for companies investing in employee development.

Enhanced Accessibility: AI tutoring provides educational and training resources across all sectors and demographics, helping to bridge gaps in access to high-quality learning and development opportunities.

Reduced Burnout: For both students and professionals, AI tutors reduce the pressure of traditional learning methods, allowing for a more balanced, customized approach to acquiring knowledge.

Support for Lifelong Learning: With AI tutors available for any stage of life, people can easily pursue continuous learning, whether advancing careers, changing fields, or developing new personal skills.

Taking the Tension Out of Big Decisions—With a Little AI Help

AI can transform how we approach difficult decisions, helping individuals gather knowledge and work through questions before sitting down with others to discuss big choices. Whether it’s a couple weighing health insurance options, families tackling a shared financial decision, or coworkers preparing for a critical meeting, AI can act as a decision consultant, allowing people to enter conversations more informed and prepared. By providing clear insights and answering nuanced questions, AI helps reduce tension, making discussions smoother and more productive.

With AI, users can explore options, get clarity on confusing terms, and even rehearse the details of an issue before talking it over with a partner, friend, or colleague. This preparation leads to more collaborative conversations, where each person feels ready to contribute, bringing knowledge rather than stress to the table.

WHY IT MATTERS

Smoother Conversations: AI helps users clarify complex topics beforehand, reducing misunderstandings and making discussions feel less contentious and more constructive.

Shared Understanding: For couples or teams, AI provides an objective look at the facts, creating a shared foundation for decision-making and reducing the chance of conflict.

Building Confidence: With the ability to research and ask AI questions privately, users can approach conversations with more confidence, feeling better equipped to discuss topics they find confusing.

Improved Collaboration at Work: In a professional setting, using AI to prepare for discussions helps everyone enter a meeting on the same page, improving teamwork and reducing miscommunication.

A New Tool for Clarity: AI supports users in breaking down jargon and clarifying terms, making it easier for everyone involved to make informed, balanced decisions without feeling overwhelmed.

Can ChatGPT’s New Search Feature Outshine Perplexity?

As ChatGPT rolls out its new search features, it’s entering the competitive ring with AI tools like Perplexity, which have already carved out a strong reputation in the search space. For users, this competition could mean more personalized, reliable, and comprehensive search experiences.

Both tools have their strengths—ChatGPT’s seamless integration with its language model offers a familiar, streamlined experience, while Perplexity provides advanced capabilities like detailed source lists, follow-up suggestions, and even citations in its API.

Each tool approaches search differently. ChatGPT search is embedded directly into the chat experience, aiming to save users time by sourcing directly from recent web content. Perplexity, on the other hand, shines in specific, detailed searches that require deeper, source-based answers.

But as AI search evolves, the real test will be whether users prefer tools that streamline the search within their AI chat experience or those that emphasize structured, multi-layered research.

WHY IT MATTERS

Streamlined Knowledge Access: Both ChatGPT and Perplexity make it easier to find answers without navigating traditional search engines, saving users time and frustration.

Quality of Information: Users can leverage ChatGPT’s embedded sources for quick insights, while Perplexity’s in-depth responses are great for complex questions and research-heavy queries.

Customizable Search Options: With Perplexity’s API providing citations and other custom features, companies can develop more tailored, AI-driven information tools.

Emerging User Preferences: As search becomes a core AI feature, user choices will shape the future of search interfaces, favoring either a simplified, embedded approach or a research-focused, citation-rich experience.

Just Jokes

AGI in 2025?

Karl shared an interview with Sam Altman from @tsarnick on X. When asked what Sam was looking forward to the most in 2025, Sam seemed to say AGI. However, Sam might have been answering the first part of that question, “What are you most excited about” and not specifically referring to 2025. In fact, he said his husband and him are expecting a child next year and that was what he was most excited about in 2025.

So is AGI closer than we think?

Only time, and probably Sam, will tell.

Microsoft is getting coy with CoPilot

Eran pointed out that he now gets notifications when he logs into his Microsoft 365 Plan that he has CoPilot.

But on second look, he is just getting a few credits to use CoPilot and would have to upgrade to a business plan in order to get more.

Considering CoPilot isn’t that great, you would think Microsoft would spend more time making it actually useful before worrying if they can get their home subscribers to upgrade to a business plan for significantly more money.

ChatGPT Desktop App Now Has Advanced Voice

Jyunmi called out that he now has advanced voice in his ChatGPT desktop app. Brian checked and has it as well on the Windows version.

We are one step closer to working at our desks being able to just randomly start talking and having our computers respond and even take action for us.

Imagine in a year or two going about your work, and then suddenly saying, “This Gantt Chart is out of date for our project. Can you update it before I meet with my client tomorrow? You can find all the info you need in our emails and our last 2 meetings.” And without prompting or even pointing to the chart, you just continue working on other tasks.

Did you know?

In the 2024 U.S. presidential election, AI was both a defender and a disruptor. Advanced AI tools helped social media platforms quickly detect and flag misleading content, stopping misinformation before it could gain traction. AI algorithms scanned posts, videos, and images for signs of manipulation, even flagging deepfakes that have grown more realistic since 2020.

But here’s the twist: the same AI technology also made it easier than ever to create convincing misinformation at lightning speed. Sophisticated generative AI tools enabled rapid production of fake news articles, manipulated videos, and images that spread across the internet before detection systems could catch up.

In a battle of truth vs. technology, AI is proving to be both a powerful ally and a tricky adversary.

This Week’s Conundrum
A difficult problem or question that doesn't have a clear or easy solution.

The AI Diversity Dilemma:

AI systems increasingly influence beliefs, behaviors, and moral values worldwide, leading to a push for diverse, culturally inclusive perspectives in AI. Developers are tasked with making these systems adaptable to different cultures and ideologies. But balancing conflicting viewpoints often results in AI that represents a "neutral" or blended perspective, potentially flattening the distinctiveness of individual cultures.

The conundrum: Does striving for universally representative AI unify society by promoting inclusivity, or does it risk creating a homogenized, inauthentic version of diversity?

Should AI be designed to respect every cultural perspective fully, even if it means reflecting divisive or fundamentally different viewpoints?

Elon Musk's Influence on AI in the New Trump Administration

With Donald Trump’s recent election, Elon Musk’s involvement has sparked discussions on AI regulation. The new administration, alongside Musk’s pro-free speech stance at xAI, may lead to fewer AI regulations and more competition with countries like China. This shift is expected to impact the development and oversight of AI technology in the U.S.

Grok API Release with Free Credits for Developers

Elon Musk’s xAI recently launched the API for its AI model, Grok. This move opens up Grok’s capabilities for exploration, aiming to build a user base and foster innovation around the new API. You get $25/mth in free credits until the EOY to test out its function calling, 128k token context and custom system prompt support.

Perplexity’s Election Information Hub

Perplexity took a risky chance and rolled out an Election Information Hub a few days before the U.S. elections designed to provide users with comprehensive data on candidates, electoral maps, and key races. The new platform relied upon the AP for live election data, Democracy Works for voting logistics info, curated sources like Ballotpedia, and built-in citations for verification of information. While other AI platforms simply refused to answer election questions or flat out got them wrong, Perplexity took the opposite approach. This tool positions Perplexity as a potential trusted source for news and information, expanding its role beyond AI search.

Perplexity CEO’s New York Times Tweet Controversy

Perplexity CEO Aravind Srinivas responded to NYT publisher AG Sulzberger on X (formerly Twitter) with: "Hey AG Sulzberger @nytimes sorry to see this. Perplexity is on standby to help ensure your essential coverage is available to all through the election. DM me anytime here". Although he aimed to offer Perplexity as a resource, the tweet was perceived by some as dismissive of the striking workers, sparking debate about AI’s role in traditional journalism. After the backlash, Srinivas attempted to clarify that "the offer was not to 'replace' journalists or engineers with AI but to provide technical infra support on a high-traffic day". However, critics noted that this was still effectively offering to replace the work of striking technical staff.

Meta’s Llama AI Approved for U.S. Government Use

Meta’s open-source Llama AI has been cleared for use by U.S. government agencies, including defense contractors. Intended to aid in logistics, cybersecurity, and planning, Llama’s open-source nature raises questions about its application in sensitive areas given previous limitations on military use.

Claude 3.5 Haiku Model Price Increase

Anthropic’s Claude 3.5 Haiku model, known for its high efficiency, saw a fourfold price increase, making it more costly than the larger Opus model. This adjustment reflects Haiku’s improved performance and positions it as a valuable tool despite the higher cost.

Desktop Apps for Claude, Perplexity, and ChatGPT Released

Desktop versions of Claude, Perplexity, and ChatGPT were recently launched, bringing enhanced functionality for users. These apps make it easier to switch between models, with features like image analysis in PDFs, creating a seamless workflow for various AI tools.

Whale Tracking and Translation with AI

Project CETI has expanded its efforts to decode whale communication by using AI and VHF sensors to predict whale surfacing locations. This technology helps researchers gather vocalization data to understand whale “language,” advancing marine biology and conservation efforts.

UC San Diego’s Ultrasound AI for Muscle Monitoring

Researchers at UC San Diego have developed an AI-driven ultrasound patch capable of monitoring muscle activity in detail. Initially aimed at respiratory illnesses, this technology can now capture fine motor movements, potentially aiding in robotics and exoskeleton control.

Amazon Prime Video’s AI Summaries and Scene Recaps

Amazon Prime Video is enhancing its X-Ray feature with AI-generated summaries and scene-specific recaps. This tool allows viewers to quickly catch up on prior episodes or specific character interactions, adding convenience for binge-watching and long breaks between seasons.

Google AI Coding in 25% of New Code

Sundar Pichai revealed that Google’s AI systems now generate 25% of the company’s new code, showcasing Google’s use of AI in software development. This shift could lead to faster feature rollouts and highlights the increasing integration of AI in tech workflows.

Did You Miss A Show Last Week?

Enjoy the replays here on YouTube or take us with you in podcast form on Apple Podcasts or Spotify.