09 January 2025

9 AI Tools You MUST Know In 2025 (Productivity, Audience Growth & More)

1Frameworks

8Insights

Frameworks in this episode

Marketing5 steps

Expertise-First AI Audience Growth

Build expertise first, then use AI to multiply proven content skills

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Hot Take· 1

Hot Take07:30

Integrated AI Platforms Threaten Standalone Point Solutions

Although Cubby provides a compelling focused experience, the host questions whether standalone research products can survive competition from platforms with native access to users' existing data. Gemini's connection to Google Drive and YouTube illustrates how integration can outweigh a specialized feature set.

Platform integrations reduce setup friction
Gemini can query material already stored in Google Drive
Large ecosystems can absorb point-solution features
Specialized tools must offer value beyond convenient retrieval

“I wonder how they survive the integration that Google has in in uh things like Google Drive”
07:30

“it does make me wonder around how these kind of Point Solutions will compete against the conglomerates like Google”
08:30

#gemini#google-drive#platforms#competition

Explainer· 1

Explainer07:00

AI Converts Scattered Sources Into Accelerated Learning

The host describes AI as a way to digest a collection of unstructured sources and retrieve the essential information about one topic. Its practical value is less about storing more material and more about reducing the effort required to locate, combine, and understand what has already been collected.

Centralizes scattered learning material
Extracts topic-specific information
Reduces manual searching across long sources
Organizes otherwise unstructured personal data

“what AI really helps you do is it's like an accelerated learning tool”
07:00

“we all are unstructured data makers like we have unstructured data everywhere”
09:00

#learning#research#unstructured-data#productivity

Story· 1

Story22:00

A Family Scheduler Shows Why Personalized Software Matters

The host used Replit's AI agent to build a side-by-side daily scheduler tailored to life with a ten-month-old child. The example supports a broader prediction that people will increasingly create narrow software for their own exact workflows instead of accepting every feature and constraint of mass-market products.

AI agents lower the barrier to building functional software
The scheduler reflects one family's exact planning needs
Bespoke tools can omit unnecessary commercial features
Personalized software may disrupt established software categories

“this really is an example of why we're all going to have personalized software”
22:30

“I don't need a tool with all of the features I just need some that are bepoke to me”
23:30

#replit#personal-software#coding#product-building

Tool· 4

Tool02:30

Flow Turns Voice Into an Interface for Everyday Work

Flow integrates with desktop applications so users can dictate, revise, and translate text without relying on a keyboard. The host sees particular value for quick Slack and email responses on mobile, although people who formulate ideas through writing may need time to adapt.

Controls writing workflows through voice
Works across applications such as Slack and Google Docs
Supports live corrections and translation
May suit quick mobile replies better than deliberate drafting

“what Flo does is actually integrates into all of your apps and you can actually write back via just your voice in AI transcribing that…”
03:00

“I do find that I have to write something out to be able to formulate it in my brain”
05:00

#voice#productivity#dictation#workflow

Tool05:30

Cubby Makes Mixed-Media Research Searchable

Cubby combines PDFs, podcasts, videos, articles, threads, and local files into a topic-specific research repository. It transcribes and indexes the material so an AI assistant can answer questions across sources or isolate selected portions of a source.

Combines many media formats in one research space
Transcribes audio and video content
Supports questions across the collected material
Can restrict analysis to selected portions of a source

“it's actually allowing you to build a context window basically for an AI agent based upon anything you want to put into a single folder”
06:30

“you can actually add in content and take the parts that you actually want to have the AI aggregate with the rest of the content”
08:30

#research#knowledge-management#transcription#ai-assistant

Tool24:00

NotebookLM Turns Source Folders Into Briefings and Study Guides

NotebookLM connects source collections to an AI workspace that can generate briefing documents, audio reviews, podcast-style discussions, and study guides. The host uses it across multiple work folders and considers its expanding feature set substantial enough to warrant a dedicated episode.

Connects AI analysis to collections of source material
Creates briefing documents
Generates audio reviews and podcast-style output
Builds study guides from supplied content

“I use notebook llm all the time”
24:00

“this study gu sty guide is really cool it basically allows you to create a study style guide on any top and any kind of…”
24:30

#notebooklm#research#study-guides#knowledge-management

Tool24:30

Napkin AI Generates Visuals for Business Content

Napkin AI differs from general image generators by producing diagrams and visuals suited to business communication. The host demonstrates using a LinkedIn post about AI-driven procurement as input and receiving multiple visual options that could accompany the post.

Focuses on business rather than consumer imagery
Transforms existing written content into visual options
Can support LinkedIn and other professional content
Produces several representations of the same business idea

“napkin AI is another image generator but very different it generates business images not uh kind of consumer images”
24:30

“this is really great for like business”
25:00

#napkin-ai#business-visuals#linkedin#image-generation

Takeaway· 1

Takeaway09:30

The Best Meeting Notes May Require an Extra Step

Dedicated note-taking applications automate meeting transcription and summarization, but the host reports better results from manually sending a Zoom transcript to ChatGPT. This creates a trade-off between the convenience of an automatic workflow and the quality of a more deliberate process.

Many tools automatically transcribe meetings
A Zoom transcript can be processed separately in ChatGPT
Manual processing may produce better notes
Automation and output quality should be evaluated independently

“it gives me the best med notes of any tool that I've used but it's a little bit of a process whereas these things do…”
10:00

#meetings#notes#chatgpt#automation