OpenTube

Apple's New Mac Line is Built Around Local AI. The Bet Is You'd Rather Own Than Rent.

AI News & Strategy Daily | Nate B Jones

350,925 views

Apple rebuilt the entire desktop Mac line around local AI, and the price ladder tells you exactly who it is for. Here is what the M6 and M5 chip split means for running local models, keeping agents alive on your desk, and deciding how much AI compute is worth owning.

My Links 🔗
👉🏻 Newsletter: https://natesnewsletter.substack.com/
👉🏻 X: https://x.com/natebjones
👉🏻 TikTok: TikTok: nate.b.jones
👉🏻 Instagram: Instagram: nate.b.jones

What's really happening inside Apple's local AI launch?
The common story is Apple coming for NVIDIA — but the real question is which intelligence is worth owning and which is worth renting.

In this video, I share the inside scoop on Apple's local AI bet and the cloud bet running against it:

Why Apple put its newest chip at the bottom of the line
How the memory ladder maps to how many agents you can run
What the cloud bet assumes that a local machine cannot answer
Where the missing middle sits and who could own it

Local compute is now a real option for a profitable minority of prosumers, but frontier agents are moving to persistent cloud machines fast enough that buying the hardware does not settle where your work ends up living.

Chapters:
00:00 apple rebuilds the entire desktop mac line around local ai
00:39 the strange wrinkle, the new chip is at the bottom
01:34 own your intelligence or rent it
02:46 the memory ladder from mac mini to mac studio
05:14 the obvious objection and the memory math videos
06:29 two trends moving in opposite directions
08:27 what these machines cost and which one i would buy
11:22 the setup problem apple has not solved
12:43 the cloud bet is a separate bet
14:33 renting a personal supercomputer instead of a chatbot
17:45 the real danger, your mac becomes a terminal
20:33 my bet on the missing middle

Listen to this video as a podcast.

Spotify: https://open.spotify.com/show/0gkFdjd...
Apple Podcasts: https://podcasts.apple.com/us/podcast...

20 Comments

@dmdunn11d 1 month ago (edited)

I bought my first McIntosh in the mid 90s. For the last 31 years I’ve been surfing the computer tech wave as a desktop publisher. My learning curve began with Aldus PageMaker and continued into the present with Adobe Creative Suite. The energy of this evolving technology has kept me on the board. In 2025; My last paid gig was designing and desktop publishing a memoir for a longtime partner, client, and friend.

I’ll turn 84 this fall. AI offers the energy of a new wave, a new learning curve, and new creative-ventures-in-retirement. The ability to ask an AI assistant “how do I do this,” gives me the safety of a smart outrigger on a new board.

@BavariaR 1 month ago

I think one of the main motivations for local AI models is confidentiality, sovereignty, and independence from large frontier models that could disappear anytime at the whim of others. If we had a setup with cloud models not owned by a single greedy company but owned jointly by humanity, and that could ensure our data is safe and the models are not biased, then many individuals would no longer care about purchasing hardware to run models locally.

@pmbrig 1 month ago (edited)

I'm not a "prosumer," I'm an ordinary bloke embedded in the Apple ecosystem. I think there are a lot of us who want AI to streamline daily tasks and have no need to bespoke crunch huge databases. I'm solidly in the local-AI/cloud-when-needed camp. I'm still suspicious about cloud models storing and accessing all my data. Now, I know that there are also lots of folks who need something more for sophisticated and more power to them, but I don't need to pay an enterprise model for monthly tokens with questionable privacy. I think you're right on target about both models doing well for different user bases.

@zebratroopfan6695 4 weeks ago

I use my computer to organize my life. I track my health, finances and schedule my daily tasks. I don't code, edit or design. I check my email. I watch a bit of YouTube. I hope I'm reminded what I'm supposed to do or be for the day. I’m in the Apple ecosystem and i like it. The new siri is exciting because it is using my personal data on my phone to make my life function a little easier. It’s that simple for me.

@luigi_monaco 1 month ago

I think the Mythos/Fable shutdown killed the trust into a cloud solution. Forever.

@coonhound_pharaoh 1 month ago

Apple is on the record stating their excitement about technology introduced in the M7 made investment in M6 get cut short. The reason the M6 is only in cheap units is because Apple is gearing up for the M7.

@keme8541 1 month ago

The slow down in frontier releases means open weights have nearly caught up. I vote for local AI, as privacy is vital for proprietary data.

@peterscheepers5780 1 month ago

I’m a retired industrial control engineer and in the controls industry we have seen the shift from centralized to distributed to the extend that sensors are even highly intelligent and still getting more affordable. Keeping the intelligence as close as possible to the place it is needed is a fail soft benefit, will be cheaper in the long run and gives you control over your own data. You don’t need a Bentley to drive to the supermarket, good enough will do. Yes I think there will be a future for frontier ai but i expect the majority of ai to be distributed locally.

@jc2044 1 month ago

The problem with cloud services is the uncertainty they introduce: service availability and reliability, data security and privacy, and the constant risk of unexpected changes to policies, pricing, or terms of service.

The number of horror stories just keeps piling up. It is difficult to ignore the risks of putting critical data and operations in someone else’s hands.

@00w4gap00 1 month ago

Absolutely local first, frontier second.

@ryanvacation7319 1 month ago

I can also imagine a lot of big companies want to keep their data private, so they most likely prefer local.

@machine1st 1 month ago

Small models are the future.

@uclassc 1 month ago

The biggest thing that keeps me from from writing agents is if they go rogue I may not know or realize too late, no guardrails

@davidcampos9768 1 month ago

I’ll definitely buy one . I want local AI as much as possible

@StephenFlournoy 1 month ago

100% worth it for local models. Easy to solve the routing issue, just use frontier model to build the router "you need". I'm already doing this sharing my compute with a friend.

@Christ-on-a-Bike 1 month ago

Local AIs for image and 3D model generation are going to be a really big thing. Because in a strange way there is kind of a finite amount of AI you need there, you are already seeing it in some of the most advanced image and even video generation models.

@vincentbutton5926 1 month ago (edited)

I was running always-on personal assistants... then the token bill came. That's solved now.

@RaymondSiu_Love_God 1 month ago

I think you’re missing one major factor: data privacy.

For non-technical users, data privacy is often a bigger concern than performance. Running an LLM on our own hardware can significantly reduce the risk of data leakage.

Our team is currently piloting Apple hardware to see whether it can run LLMs locally while still delivering results comparable to frontier models.

@bitphr3ak 1 month ago

The mainframe didn't disappear when the PC arrived, both do different things, the AI will diversify.

Both, I'll use both

@kenlewis2253 1 month ago

Prosumers will always be gearheads. There’s something inherently interesting in the tactile and in ownership. That’s not going away.

Everybody else will prefer cloud because it is as easy to access as opening a website