Dconstructed: The Podcast

Tony and Matt break down a heavy week of earnings and tech news:

Prepping and doom talk: Tony explains why he's stocking up on supplies.

Microsoft, Alphabet and Amazon earnings: strong cloud growth, but AI spending is pushing cash flow negative. Microsoft also offered buyouts to about 60,000 employees.

Intel: shares fell about 25% after weak guidance.

Anthropic's S-1: the IPO plans, the loss figure and the special voting share.

Apple: the new touch-screen OLED MacBook Pros, plus iPhone Duo hinge complaints.

Hands-on reactions: Matt tries the iPhone Duo and Meta's display glasses at Best Buy.

Not financial advice.

Episode 124 is here, and earnings season is in full swing. Tony and Matt break down the first reports of the season:

Netflix: a profit beat, but weaker guidance and a lowered ad revenue forecast sent the stock down. They also cover layoffs and the Narnia sequel cancellations.

Meta: revenue up 26% to $64.6B, but heavy AI spending left free cash flow negative.

SpaceX & Tesla: SpaceX's AI business drives big growth. Tesla beats on cash flow and cuts its capex guidance.

Tesla's flying Roadster: a real demo or just a gimmick?

iPhone Duo: early reviews, screen complaints and why it's still sold out.

Not financial advice.

Matt and Tony unpack OpenAI's DevDay announcements and debate where always-on AI assistants fit into work and everyday life. They compare voice workflows, model choices, and subscription pricing, then discuss Tony's Tuppence app launch, App Store review frustrations, and Matt's new Mac Mini setup.

Takeaways

• Useful agents need clear boundaries between personal and company information.
• Voice can be convenient on a commute, while shared offices present different challenges.
• Model choice and subscription value depend on the work you actually need done.
• Launching an app means navigating reviews and understanding how families will use it.
• An always-on Mac Mini opens up new possibilities for coding and assistant workflows.

Chapters

00:00 Welcome to Episode 122
01:12 OpenAI DevDay and the Dots rollout
02:44 Personal agents versus workplace boundaries
06:47 Voice assistants and calling your AI
09:34 Choosing AI models for everyday work
11:59 AI IPOs and market uncertainty
14:46 AI subscription pricing and subsidies
16:57 Codex Cloud, Spaces, and Pages
20:24 Plugin extensions and the ChatGPT ecosystem
24:21 Tuppence launches: chores and allowances
28:02 App Store review frustrations
29:22 Setting up an always-on Mac Mini
32:34 Wrap-up

Follow Dconstructed Podcast: https://www.youtube.com/@DconstructedPodcast

Matt and Tony unpack NVIDIA's earnings and the AI spending boom, discuss Hugging Face and local models, and weigh Apple's hardware lineup. Matt debates the Mac Mini specs he needs for development and an always-on AI assistant.

Takeaways

• Strong AI demand still leaves questions about spending, financing, and long-term returns.
• Local models offer more control, but the hardware and setup need to fit a real use case.
• Memory and storage choices matter when buying a Mac for development and local AI.
• An always-on home machine could bring coding, automation, and personal storage together.

Chapters

00:00 Welcome to Episode 118
00:51 NVIDIA earnings and AI demand
02:56 Circular financing and the AI spending boom
08:13 Hugging Face and NVIDIA's AI ecosystem
12:28 Why run AI models locally?
15:31 OpenAI, Cursor, and model-provider dependence
16:28 Apple's event and foldable-phone expectations
20:32 Mac Mini and Mac Studio: choosing the right specs
26:26 ChatGPT for personal finances
30:18 An always-on Mac and the wrap-up

Follow Dconstructed Podcast: https://www.youtube.com/@DconstructedPodcast

Matt and Tony discuss Meta Connect's glasses and wearable AI ambitions, compare always-on assistants, and share their latest coding-agent experiences. Matt also recaps a GrokBot meetup and the gap between promising demos and useful everyday workflows.

Takeaways

• Wearable AI has to balance comfort, price, usefulness, and privacy.
• Camera-free glasses may address some concerns while leaving others unresolved.
• Agent-to-agent collaboration can help reuse knowledge across development projects.
• Always-on assistants still need clear use cases and trustworthy access to personal accounts.

Chapters

00:00 Welcome to Episode 121
01:02 Meta Connect and VR glasses
06:45 How might Apple respond?
07:31 Camera-free Ray-Bans and privacy
11:10 Muse Charm and wearable AI
12:39 The always-on assistant landscape
16:45 Opus and coding-agent collaboration
20:15 GrokBot and the engineering meetup
22:45 Agent workflows, credentials, and local alternatives
24:55 What's next for assistants and wearables
25:35 Wrap-up and the podcast milestone

Follow Dconstructed Podcast: https://www.youtube.com/@DconstructedPodcast

An App Store rejection sends Tony back to his referral system. Matt and Tony compare subscription workflows, talk inflation and markets, and explore the growing field of always-on AI assistants—from Meta Muse and GrokBot to an SMS-based assistant.

Takeaways

• Referral rewards need to work with the app's subscription and payment flow.
• AI-assisted coding still requires product judgment and an understanding of edge cases.
• Always-on assistants differ in interface, integrations, and how much account access they need.
• A focused task such as monitoring travel prices can make an assistant's value easier to judge.

Chapters

00:00 Welcome to Episode 120
00:48 App Store rejection and referral rewards
05:45 Referral codes, links, and subscription edge cases
10:11 Vibe coding still takes judgment
11:16 Inflation, the Fed, and markets
15:07 Housing, rates, and hardware prices
19:04 Meta Muse and always-on agents
21:46 GrokBot and privacy trade-offs
23:42 An SMS assistant for travel planning
31:40 Wrap-up and the changing season

Follow Dconstructed Podcast: https://www.youtube.com/@DconstructedPodcast

Building an app is only half the battle. Matt and Tony talk AI side projects, App Store launches, and the challenge of finding users—plus Apple's home-device rumors and SpaceX's mobile ambitions.

Takeaways

• Shipping an app doesn't solve distribution: reaching the right audience takes a separate plan.
• AI can help with promotion, ad monitoring, and support workflows, with human review at key steps.
• Useful niches and a better experience can matter more than adding another generic app.
• Agent workflows need clear inputs and boundaries; they don't always need to run continuously.

Chapters

00:00 Welcome to Episode 123
00:44 Apple's surprise event and smart-home rumors
04:48 SpaceX, spectrum, and Starlink's mobile ambitions
10:31 Tony's apps reach the App Store
12:02 The distribution problem and AI-assisted promotion
15:09 A metronome app for students and musicians
17:09 Matt's HVAC expansion and finding a niche
18:59 AI agents, ad monitoring, and support workflows
21:40 Competitive research and the Leeway experiment
22:47 Tony's trading-bot experiment
24:01 Wrap-up and what's next

Follow Dconstructed Podcast: https://www.youtube.com/@DconstructedPodcast

We cover the government shutdown of Anthropic's new Fable model, go in depth on SpaceX's public listing, and reflect on Apple's WWDC.

Takeaways

• App development and the challenges of using Fable for coding.
• The impact of recent events on the stock market and the celebration of the Knicks' championship win.

Chapters

• 00:00 Introduction and Agenda
• 05:15 Cancellation of Fable
• 10:33 IPOs and Stock Market
• 18:02 Apple's WWDC and Siri AI
• 27:38 Knicks' Championship Win and World Cup

We dive into the wild world of AI with a lively chat about Claude Design's game-changing impact on product design. We're talking big moves by Google and Tesla, and the epic showdown in the AI arena. It's all about saving time, sparking creativity, and having a blast while doing it!

1. Intro - 0:00
2. Claude Design Buzz - 5:00
3. Big Tech's Big Bets - 15:00
4. AI Showdown - 25:00
5. Design Opportunities - 35:00
6. Wrap-Up - 45:00

We talk about the AI shopping accusation against Perplexity and the company's recent appearance in the news. Plus, Amazon's AI investments are formally starting to pay off.

Takeaways

• AI shopping accusation
• Perplexity in the news

Chapters

• 00:00 AI Shopping Accusation Against Perplexity