Justas Petronis
banner
dialethics.io
Justas Petronis
@dialethics.io
Doctoral student in Philosophy of AI | Product @ theydo.com | from Vilnius, Lithuania | my substack is dialethics.io
The Sea — a symbol of Lithuanian Cultural Movement against the fascistic party Dawn of Nemunas taking over the cultural sector in our new government youtu.be/3kGTf5yXW1g?...
The Sea
YouTube video by Lithuanian National Symphony Orchestra - Topic
youtu.be
October 5, 2025 at 10:12 AM
Paternity leave.
August 24, 2025 at 4:57 PM
Creature 3.
August 1, 2025 at 4:51 AM
Lost for words how good this one is. What’s trust, what’s an internalized, persistent belief, and what’s pretense, charade, and a ruse? And what’s the point of it all — when you’re, supposedly, not that different from the other guy. All at a backdrop of a fallen empire longing for a raison d’etre.
May 3, 2025 at 11:07 AM
Creature 2.
April 12, 2025 at 3:13 PM
April 12, 2025 at 3:11 PM
Creature.
April 2, 2025 at 7:19 PM
Trump’s getting peer-reviewed by his clique now in open source publications. What a time
March 28, 2025 at 11:51 AM
We weep for US. Sincerely, Europe.
The US Director of National Intelligence is retweeting a far-right conspiracy theorist based in Malaysia.
March 18, 2025 at 5:15 AM
Happy 35th, my lovely 🇱🇹.

As my mom said today, maybe there were hardships along the way but so far — the happiest stretch of history in a very long time.

Chaos globally can’t take away from that pride, especially being where we belong — 🇪🇺.
March 11, 2025 at 1:02 PM
Reposted by Justas Petronis
I often hear that we need realism, not idealism. But realism never brought anyone peace and never liberated any oppressed nations. Only the dogged determination to defend our values ever achieved the victories we cherish. Only dreams and hard work can make us great again.🧵17/17
February 22, 2025 at 9:10 PM
Reposted by Justas Petronis
The sounds of war in the sky of Kyiv right now
February 20, 2025 at 8:41 PM
Reposted by Justas Petronis
Kamala was right. And it only took him less than 30 days to roll over for Putin.
February 19, 2025 at 7:52 PM
I genuinely share Mr Heusgen’s sentiment. New world order. www.rte.ie/video/id/232...
Munich Security Conference chairman sheds tears
www.rte.ie
February 17, 2025 at 4:25 AM
It seems to me that some Europeans are now starting to feel viscerally the existential dread, broken trust, and betrayal that Ukrainians have lived with for over a decade. www.rte.ie/video/id/232...
Munich Security Conference chairman sheds tears
www.rte.ie
February 17, 2025 at 3:35 AM
We are embedding AI in our moral decisions. Are we, then, thinking more clearly—or just outsourcing our moral agency to synthetic systems? I try to unpack this super high-level, as if this was the intro to my thesis: www.dialethics.io/p/the-burden...
The Burden of Infinite Memory
An attempt at an introduction for my thesis while preparing a conference presentation in Tallinn this March (2025)
www.dialethics.io
February 3, 2025 at 8:04 PM
Reposted by Justas Petronis
Historian of fascism here. It was a Nazi salute and a very belligerent one too.
January 20, 2025 at 10:16 PM
That was some damn fine cup of coffee, Mr Lynch 😢
January 16, 2025 at 7:12 PM
We will never forget. #jan13 youtu.be/io6PTL81ocQ
January Events - How Lithuanian people stood up to Russian tanks in 1991
YouTube video by The Soup Central
youtu.be
January 12, 2025 at 8:28 AM
Reposted by Justas Petronis
can someone pls tell me where these superhuman models are
December 31, 2024 at 2:08 PM
Munch Museum Oslo
December 27, 2024 at 2:24 PM
The machines didn't fail to think like us. We failed to understand how we think.

When we taught machines to recognize patterns instead of following rules, they began to learn like humans – and in doing so, revealed our fundamental misunderstanding of our own minds. www.dialethics.io/p/system-error
System Error
Why We Can't Compute Ethics
www.dialethics.io
December 4, 2024 at 6:11 AM
Why the analogy between AI and atom bomb is not correct?

- AI's proliferation is fundamentally different (runs on commodity hardware)
- The attack surface is vastly larger (every connected device)
- The development cycle is orders of magnitude faster
- The barriers to entry are significantly lower
November 19, 2024 at 10:21 PM
A colleague of mine responded to my previous text: "with closed source, you can only trust, not verify". I admit, it's a compelling argument for open source AI. But what happens when the ability to verify doesn't translate into actual verification?

www.dialethics.io/p/the-glass-...
The Glass House Fallacy
Part II of my challenge that open source (AI) transparency in itself does not lead to better, more ethical development outcomes
www.dialethics.io
November 19, 2024 at 7:40 AM
Musing about the naiveté of open source AI development tomorrow. Find it here in just a few hours: open.substack.com/pub/petronis...
November 17, 2024 at 6:00 PM