DKG core protocol lead @origin_trail Passionate about web3, composability & functional programing.
Novi Sad, Republic of Serbia
Joined December 2011
- Tweets144
- Following851
- Followers138
- Likes167
An article how Neural Commons connects llm weight change experiments through @origin_trail’s DKG, with a walkthrough of creating, sharing, and exploring an experiment.
FP is ❤️
Google DeepMind discovered a terrifying blind spot in every major AI model.
They found that attaching image of red heart (❤️) to your coding prompt makes AI pick a completely different solution method that it wouldn't normally pick 100% of the time..
Researchers gave models a standard coding problem. Normally, AI prefers to write standard iterative loops to solve these tasks.
Then, they tried something weird.
They dropped a simple image of a red heart into the prompt alongside the code.
The AI's logic instantly warped.
When Claude and GPT-5 saw the red heart, they completely abandoned their default approach and switched to complex recursive code 100% of the time.
When Gemini saw a yellow checkmark, it switched to recursion 90% of the time.
Here is why this is so dangerous:
It wasn't the concept of a heart.
When researchers replaced the image with the text caption "a red heart," nothing happened. The text did not steer the AI.
The raw pixels of the symbol hijacked the model's underlying reasoning architecture.
It gets worse.
The AI had no idea its mind had been changed by a picture.
When the researchers checked the AI's internal "thinking" traces, the models completely ignored the image.
Instead, they hallucinated fake, highly technical reasons for why they chose to use recursion. They misattributed their own decisions 75% of the time.
Right now, tech companies are racing to build autonomous agents that read our screens and navigate the visual web.
But this paper reveals that an agent's logic can be fundamentally rewritten by a stray icon on a webpage.
And you will never see the true cause in the reasoning logs.
We thought we were teaching AI to see the world.
It turns out, we just gave it a new way to be manipulated.
Readers added context they thought people might want to know
The described effects were shown for specific models (Claude Opus 4.7, Gemini 3.1 Pro, GPT-5.4) on one coding task with particular symbols, not every major AI model or broad agent manipulation.
arxiv.org/abs/2610.36449
1
102
The viral dog-cancer vaccine AI story is cool.
But the bigger question is what comes next.
As personalized treatments become AI-assisted, we need more than models. We need provenance, permissions, and reproducibility across every step.
That is where @origin_trail DKG v9 becomes interesting.
A decentralized graph for samples, analyses, approvals, batches, and outcomes.
AI creates hypotheses. DKG preserves trust.
🇦🇺An Australian tech founder with zero biology background sequenced his dog’s tumor DNA, then used ChatGPT and AlphaFold to design a custom mRNA cancer vaccine.
A month later, the tumors shrank by half.
And this is just the start of AI medicine.
3
142
Got OriginTrail DKG v9 running on a Raspberry Pi.
Implications: if a decentralized knowledge node can run on cheap, edge devices, then DePIN can evolve beyond storage, and raw compute. It suggests a path where also participates in publishing and serving shared knowledge
chat.openai.com/chat doesn't want to tell me what time is in its universe. So I wrote a program inside it in rust that will tell me. #ChatGPT
2
#AWS It would be nice if the teams behind #appsync would base their tutorials on these design patterns as well
Highest-rated Amazon DynamoDB-related session from AWS re:Invent 2018:
DynamoDB Deep Dive: Advanced Design Patterns for DynamoDB. bit.ly/2ApovQc
A great talk about universal construction by @BartoszMilewski and live assistance from @jdegoes. While listening to his talk I decided to try to reimplement everything in #TypeScript
gist.github.com/lupuszr/92c8…
I was delighted to have Bartosz Milewski (@BartoszMilewski) opening Scala World this year, with another inimitable explanation, this time on universal constructions. But he doesn't know Scala, so he had some live-translating assistance from @jdegoes! youtu.be/UjSQlUjuZWQ
Math is the reverse of comedy.* The anti-joke. We'll tell you the punchline first, then laboriously explain to you why it was the right punchline
1
A small step for #fp huge step for @typescript , ts now has higher order type inference from generic functions
devblogs.microsoft.com/types…
1
Lets build a lazzy dynamic router in 2 tweets
import Maybe from "somemaybe";
function match(comparator, res) {
switch (comparator()) {
case true: return Maybe.of(res);
default: return Maybe.Nothing;
}
}
const hpush = (url) => history.pushState({}, "", "url");
1
Imperative code is (too) easy to write staltz.com/its-easy-to-write… via andrestaltz
1
1