Pinned Tweet
The next billion software downloads may not be initiated by humans: linkedin.com/pulse/ai-agents…
🤖 Made with AI
1
2
208
Yep. @Uptodown received 3,600+ new apps last month. AI has opened the floodgates.
Here’s what gets through:
reddit.com/r/uptodown/commen…
Open source built the internet. It deserves more than a footnote.
We’re giving it more visibility at Uptodown, starting with clearer open-source license details on app pages.
More to come 👇
1
2
130
Uptodown now shows more open-source license details on app pages. Know what you can use, study, modify, and share: reddit.com/r/uptodown/commen…
34
Listen, @OpenAI, @AnthropicAI. Once ARC-3 or FrontierMath are saturated, the AGI problem may no longer be about depth. It may be about everything around it:
Real two-way communication.
Continuous listening with attention, the ability to realize when something is meant for you.
And non-reactive behavior: reasoning that doesn't begin when a prompt arrives and end when the answer is sent, but runs continuously. What we call our inner dialogue.
If the labs connect those three pieces, you'll have pushed the boundary of what it means to be human.
And we can't rule out the possibility that there may be nothing left behind that boundary.
3
197
No puede ser que un día estés cambiando pañales y al siguiente haciendo sus maletas para mudarse a @UWMadEngr 🥲
🤖 Made with AI
9
31
1,603
When Software Starts Installing Software. AI agents have entered the app store. Unfortunately they refuse to click Download :) linkedin.com/posts/luishgcom…
106
Who Downloads Software When Humans Stop Clicking?. The Download Button Was Never the Important Part:
luishg.com/2026/08/27/if-you…
1
218
Uptodown Desktop for Windows and macOS in 2 minutes. Your computer. Your apps. Your rules :)
en.uptodown.com/marketplace/…
3
153
Privacy
Is
A
Human
Right.
Want to download software anonymously?. No account required, no personal data handed over, no selling your soul to the devil.
Uptodown Desktop for Windows and macOS:
en.uptodown.com/marketplace/…
🤖 Made with AI
2
99
Dear Windows and Mac users: welcome to software freedom. Uptodown Desktop 1.0 is now available for you: en.uptodown.com/marketplace/…
Hey @ollama, biggest pain using your cloud in AI agent frameworks (@openclaw @NousResearch ) is that there’s no endpoint to check usage limits.
Minimax M3 is a beast to run on your service, but please give us a way to check this without hacking ollama.com/settings 🙏
102
Forcing identity into online services doesn’t create safety. It creates control infrastructure. And once it exists, it never stays limited to its original purpose. Systems don’t care about your ideology.
It is for parents to raise their children. Not platforms.
The European Age Verification App is ready ↓ teapaw.org/i/broadcasts/1lJQRvjoX…
Readers added context they thought people might want to know
Russia started internet regulation for exactly the same reason in 2011. Then, gradually, blocklists, VPN bans, Sovereign Internet law, and eventually whitelists in 2026.
loc.gov/item/global-le…
1
1
98
Your favorite private, local AI-powered open-source terminal for Linux, Windows, and Mac “open os cli” now supports reasoning mode and can use tools like the browser: github.com/luishg/open-os-cl…
108
I like the idea of defining computers as "non-meat-based entities."
Replying to @Hesamation
Until human consciousness is understood, denying the possibility of non-meat-based consciousness is like trying to drive in a nail with a hammer made of jello.
48
Windows 10 just entered our OS Sovereignty Index and you are NOT ready. Windows 10 scored 52, Android 41, Windows 11 39. Microsoft killed Windows 10, it still respects you more than the phone in your pocket, AND more than its own successor index.open-os.com/
103
Team A just got a new seat: @MurenaCom's /e/OS drops into the Open-OS Index with an 84/100.
De-Googled by default, microG where Google used to be, 287 supported phones. Pretty wild for a "side project" of a French non-profit index.open-os.com
1
110
Just shipped index.open-os.com. The Freedom Index for Operating Systems. 9 factors, 3 axes, every score evidenced.
▎@GrapheneOS @fedora → A. Hats off.
▎ @android → C. C'mon, y'all can do better.
▎ iOS → F.
▎ Rubric + data on GitHub
Damn, wtf is up with AI guardrails. I can't even get Gemma 4 to code "Global Thermonuclear War" from WarGames in Python because it considers it a realistic enough use case that someone could use that info to maximize damage :O
60



















