I have a problem with AI news: too much of it, and most of it arrives as a link in someone's DM three days after it happened. I'm an IT admin, and my job is already full, so I stopped trying to keep up everywhere and narrowed it down to one feed I check for ten minutes in the morning.
That feed is the news desk on AIMarketly, which pulls fresh model releases and product news from MIT Tech Review, TechCrunch AI, and The Verge into a single page. It's not my favorite thing about the site (that's the skill catalog), but it solved a real annoyance: I used to follow four sources separately, each with its own noise level, and still miss things.
What the feed actually gives you
It's dated entries with source attribution. For example, this week: the Pentagon's $30M budget line for an AI lie detector (MIT Tech Review), Lightspeed raising a $250M India fund focused on early-stage AI (TechCrunch), and Google shipping Gemini 3.8 Live with real-time avatars (The Verge). Each entry is one or two sentences, so you can scan in minutes instead of opening tabs.
For me the value isn't discovery. It's triage. I read the headline, decide if it touches my stack (local LLMs, Docker, home lab), and move on. The Pentagon story doesn't touch my stack; the Gemini avatar thing does, because Live Avatar is exactly the kind of feature that eventually lands in open models and changes what "good enough locally" means.
Why one curated feed beats four subscriptions
Subscriptions optimize for volume; every source wants you to read everything it publishes. A single aggregated desk optimizes for scanning: same stories, less repetition, no algorithm deciding your mood for the day. I don't need more AI news; I need fewer decisions about which tab to open.
The site itself is free and human-reviewed weekly (11k+ skills across 34 categories if you want to dig past the news), so there's nothing to sign up for: https://www.aimarketly.com
The honest caveats
It's a digest, not a substitute. When something matters I still read the original article at the source. And "curated" means someone decided what made the cut; if your niche is narrow (I'm in home-lab territory), you'll occasionally miss things that only landed on a smaller blog. Ten minutes of scanning plus one deep-read beats an hour of doomscrolling, though. That's the whole trade.
FAQ
Is it just RSS aggregation?
Functionally close, but with source attribution and human curation on top. The skill catalog is where it earns its keep for me; the news desk is the daily habit.
Does it cover model releases specifically?
Yes. Fresh model releases are a core section, alongside research and product news pulled from the live catalog.
Is there a cost or account?
No. Free to browse, no signup required.
Top comments (0)