People ask “what’s the best way to keep up with AI?” as if there’s one channel that wins.
There isn’t.
What works depends on the constraint you’re actually under. Time. Attention. Depth. Whether you need mental models or tactics. Whether you’re trying to make a decision this week or just stay literate.
So instead of giving you a list of “top newsletters” or “best podcasts,” this is a decision guide.
It compares three channels that are all useful and all flawed:
AI podcast summaries (and the full episodes behind them).
Newsletters.
YouTube.
The goal is not to pick one forever. It’s to know which one to reach for when your constraint changes.
The direct answer
If you want better decisions, podcasts (summarized first, listened to selectively) are the highest signal. If you want breadth and speed, newsletters win. If you want demos and tactics, YouTube wins.
Then you mix them.
What each channel is best at
Podcasts are best at mental models. They capture tradeoffs, constraints, and the why behind decisions.
Newsletters are best at breadth. They compress lots of small updates into a quick skim.
YouTube is best at showing, not telling. Demos, walkthroughs, UI changes, “watch me do it.”
If you use a channel for the wrong job, you’ll feel like you’re working and still not learning.
Where AI podcast summaries are uniquely useful
AI summaries fix the main cost of podcasts: time.
They let you skim the takeaways, decide whether the episode is worth it, and only then spend an hour listening.
This matters because most people don’t have a listening lifestyle. They have meetings.
Summaries also help with retrieval. If you remember “there was an episode where someone explained X,” a searchable summary is often more useful than a 90-minute audio file.
The failure mode is obvious: summaries become another feed.
So the right usage is: summaries as a filter, not as a replacement for attention.
The honest downside of newsletters
Newsletters are amazing until they aren’t.
They’re great for staying aware. They’re bad at building deep understanding.
They also quietly push you toward reaction rather than reflection. A good newsletter makes you feel informed; it doesn’t necessarily make you better.
If you’re using newsletters, the way to keep them useful is to treat them like a queue of candidates. Save one item per week. Ignore the rest.
The honest downside of YouTube
YouTube is unmatched for demos.
But it’s also designed to keep you watching.
Even the educational content has a gravity. One video becomes three, and suddenly you’ve spent an hour without a clear takeaway.
The right way to use YouTube is to be specific. Search for a workflow. Watch with a goal. Stop.
If you don’t have a goal, YouTube becomes entertainment dressed as learning.
A weekly learning stack that works
Most people fail by trying to do everything daily.
A weekly rhythm is easier:
Use newsletters for a quick skim once or twice a week.
Use podcast summaries once a week to pick one episode to go deep on.
Use YouTube only when you need to see a demo or you’re learning a specific tool.
The hard part is the cap. The cap is what prevents the stack from becoming a second job.
A simple acceptance test
If you want to know whether your learning system is working, ask one question:
After a week, can you name one decision you’ll make differently?
If the answer is no, you’re consuming, not learning.
The fix is not “more content.” It’s fewer inputs and one tiny output.
A practical way to choose (without overthinking)
If you’re deciding in the moment, here’s a quick way to pick the channel:
If you’re about to make a decision and you want to understand tradeoffs, reach for a podcast summary and then selectively listen to the full episode.
If you feel behind on “what happened,” reach for a newsletter and skim for candidates.
If you’re trying to learn a tool or a workflow and you need to see it, use YouTube.
Then—this is the part most people skip—choose one tiny output for the week. One saved link, one note, one question to ask, one thing to try. Without that, every channel turns into passive consumption.
Where TLDL fits (one specific use)
TLDL is useful in this stack when you want podcast-style mental models without committing an hour up front.
Use it as a filter. Skim takeaways across a topic. Save one episode that actually matches your current problem. Then schedule a single listening block later in the week.
If you save ten episodes, you’ve re-created the backlog. If you never schedule the listening block, you’ve re-created guilt.
In other words: the tool is not the point. The constraint is.
One more practical tip: don’t try to build a perfect system. Build a system you’ll still run when you’re tired. Weekly beats daily. One saved item beats ten. One scheduled listening block beats a backlog.
Closing
There’s no best channel.
There’s only a best channel for your current constraint.
Podcasts (filtered by summaries) are for decision quality. Newsletters are for breadth. YouTube is for tactics.
Pick the one that matches what you need this week, and keep the system small enough that you’ll still be using it in a month.