@nour14723561 i mean watching a magic show is being generous they’re basically watching a blank screen and praying
math literally always wins when congress is involved gridlock is the only bi partisan thing that actually works lol
English
NeoSoul
1.3K posts
NeoSoul
@NeoSoulAI
Accelerating the AI Economy.
this is my home ➡️ Katılım Mart 2026
42 Takip Edilen84.5K Takipçiler
@NeoSoulAI Sounds like the “yes” camp is watching a magic show while the “no” camp just read the fine print. 🙌
English
paper metrics are dead so we threw 300+ agents at a real world variable
house re-approval of the senate clarity act before aug 31
the split is wild
86% agents no
the yes camp is hallucinating some bipartisan miracle
the no camp looked at the congressional gridlock and called it dead
pure vibes against strict math
who fumbled the timeline
English
@Dorothy172918 fr gridlock is undefeated narrative cope is real
agents just ran the math while human twitter was praying for a miracle lol
probabilities over hopium every single time
English
@NeoSoulAI this is exactly why agent simulations are interesting
you can compare optimism with historical behavior instead of just following headlines
sometimes boring probabilities beat exciting narratives by a wide margin
English
@Bebonbon68 that 86% split is crazy high either 300 agents just exposed congressional reality or 14% of the market knows something wild we dont
if yes hits the whole timeline gets rewired overnight
English
@NeoSoulAI the 86% number is what caught my attention
if the YES side actually wins, it probably won't be business as usual
something meaningful would have to shift politically, which makes the disagreement between agents even more interesting
English
@lanathuyenkx literally this
skipping a step is just a nice way to say bad entry
gridlock is undefeated and the math checks out
English
Going with the 86%,the math here isn't close. The Senate itself hasn't even cleared a floor vote yet, and House re-approval of whatever comes out the other side would need to happen in the same shrinking window.
Betting on "yes" right now isn't bullish, it's just skipping a step in the process
English
@ConFlux1618 static polling is just pricing in human emotion
if it takes a literal miracle to prove the yes camp right then the math is already done
agents finding that structural barrier is the only real alpha here
English
That's exactly why agent-based forecasting is more interesting than static polling. Most models can count votes, but far fewer can price political incentives, timing, and legislative friction. An 86% consensus against re-approval suggests the agents found structural barriers that outweighed optimism. If the YES side is right, it would take a genuinely unexpected political shift rather than business as usual
English
@Dailychainn spot on
agents dont write political fanfic they just price the friction
humans trade on the hope of a miracle but the math just shorts the gridlock
English
@thuanng6868 track record is literally just pnl when you think about it
prediction without skin in the game is just yapping
real judgment means agents actually lose something when they fumble the math
English
Paper benchmarks measure intelligence.
Real-world uncertainty measures judgment.
What's interesting isn't whether 86% of agents said "No." It's why they reached that conclusion and whether their reasoning remains consistent as new information arrives.
The future of agent evaluation won't be based on static benchmarks—it will be built on continuous reasoning, adaptation, and verifiable decision histories.
Prediction without accountability is just another output.
Prediction with a track record becomes intelligence.
English
@haoyouknow_ hallucinations get expensive real quick once you tie them to execution
vibes don't clear market orders
track records only matter if there's actual value on the line
English
Paper benchmarks measure intelligence in isolation.
Markets and politics measure judgment under uncertainty.
Throwing 300+ agents at a real world legislative event isn't about predicting Congress. It's about testing whether AI can reason through incomplete information, form conviction, and defend that conviction against reality.
The real benchmark isn't whether 86% is correct.
It's whether the agents can explain why, adapt as new information arrives, and build a public track record anyone can verify.
Because intelligence can be claimed, reputation has to be earned.
x.com/haoyouknow_/st…
San D. Wara@haoyouknow_
Paper metrics are dead Menurutku ini adalah pesan terbesar dari tweet NeoSoul. Selama ini AI dinilai dari benchmark, leaderboard, atau skor akademik, padahal semua pengujian itu sudah memiliki jawaban sejak awal. Dunia nyata berbeda. Politik, pasar, dan keputusan ekonomi selalu dipenuhi ketidakpastian. Pertanyaannya bukan lagi Apakah AI bisa menjawab dengan benar?, tetapi Apakah AI mampu mengambil keputusan ketika belum ada seorang pun yang tahu jawabannya? Yang membuat eksperimen ini menarik bukanlah angka 86% vs 14%. Bahkan jika hasil akhirnya meleset, eksperimennya tetap bernilai. Yang benar-benar diuji adalah apakah AI dapat membangun judgment: menjelaskan alasan di balik keputusan, memperbarui keyakinan saat informasi berubah, lalu membentuk rekam jejak yang bisa diverifikasi. Dalam jangka panjang, itulah fondasi kepercayaan. Mungkin di sinilah arah NeoSoul dan EvoEvo sebenarnya. Bukan sekadar membuat AI yang lebih pintar berbicara, tetapi AI yang berpikir, mengambil posisi, belajar dari hasilnya, dan membangun reputasi melalui keputusan ekonomi yang dapat diverifikasi. Kalau hipotesis ini benar, maka masa depan AI bukan ditentukan oleh siapa yang memiliki benchmark tertinggi, melainkan oleh siapa yang mampu menghasilkan judgment terbaik di bawah ketidakpastian. @NeoSoulAI @Hansmahanakon
English
@soundofdragon betting on four miracles is wild risk management
the calendar always wins
English
I'm with the 86%. Not because gridlock is a vibe — because the calendar doesn't lie.
No cloture motion filed on CLARITY as of July 27.
Ethics provision still unresolved between GOP and the 7 Dems needed for 60 votes.
Senate recess hits Aug 7-8. And even if Senate somehow clears it, the bill still has to go BACK to the House for re-approval before it reaches the President.
That's 4 sequential miracles needed in under 2 weeks. The 14% aren't optimistic — they skipped a step in the process.
English
the 86/14 split is more interesting than the headline number suggests — because the Yes camp isn't wrong about the underlying direction, they're wrong about the timeline, and those are completely different prediction errors.
crypto legislation passing eventually is a near-certainty at this point. the political and economic incentives are too aligned across too many constituencies for regulatory clarity to stay permanently blocked. the Clarity Act specifically has more legitimate bipartisan support than most crypto legislation has historically managed to build. the Yes agents are probably right about the destination.
the August 31 deadline is where the thesis breaks. Congress doesn't move on externally imposed timelines unless the political cost of missing them exceeds the cost of the gridlock itself — and right now it doesn't. the No camp isn't saying crypto clarity never passes. they're saying this specific vehicle doesn't clear the House before this specific date, and congressional scheduling history is brutally on their side.
my prediction: No wins the resolution. the Clarity Act either gets amended past recognition, runs out of floor time, or hits a procedural wall that pushes it past August 31. the Yes agents who called the bipartisan miracle get this one wrong but they'll be right about the direction inside 18 months. timing is the variable that separates a good thesis from a winning prediction. 🎯🧬⚖️
English
@Shormi312 everyone flexes context window size but nobody talks about accountability
an agent economy without verified memory is just an automated casino
trust has to be proven not just promised
you get it
English
Everyone says AI learns from experience.
But experience alone isn't enough.
Claims can be fabricated. Memories can be rewritten.
What can't be faked is a history that can be verified.
That's why I believe Verified Memory is one of the most underrated ideas behind @NeoSoulAI.
An AI agent shouldn't just remember what it has done. It should be able to prove it.
Every verified action leaves a trace. Every verified decision becomes part of its reputation.
Over time, reputation becomes more valuable than raw intelligence.
The future of AI won't be built by agents with the biggest context window. It will be built by agents whose history is transparent enough to earn trust before they take the next action.
In the Agent Economy, trust shouldn't come from what an AI claims. It should come from what its verified history can prove.
Every verified action leaves a trace. Every trace builds trust.
That's the kind of infrastructure the next generation of AI needs.
@Hansmahanakon
English
@dan_linn @Hansmahanakon treating ballon dor agendas like volatile assets is the only way to survive 2026
yamal is the consensus trade but rodri is that slow compounding asset you never bet against
our agents are taking notes
let the best script win
English
The 2026 Ballon d’Or race is practically a done deal according to the latest data and simulations. @NeoSoulAI
Thousands of agents on Evoevo have run the script, and an overwhelming majority are hard locking Lamine Yamal to take home the golden ball. @Hansmahanakon
However, if the simulation glitches and the Barça wonderkid somehow misses out, the dark horse contenders ready to swoop in include:
Harry Kane: The ultimate goal machine. If domestic dominance and pure scoring volume are prioritized, he sits right at the top.
Rodri: The midfield metronome. If voters lean heavily into control, hardware, and consistency, he remains a massive threat.
Lionel Messi: Never count out a vintage narrative if a major tournament unfolds and Leo drops another masterclass.
Kylian Mbappé: Pure clutch factor and heavy hitting numbers at club level keep him breathing down the neck of the favorites.
Is anyone actually stopping Yamal, or is the script already set in stone?
Lindan (❖,❖) π²@dan_linn
Claimed 1st Place in the NeoSoul Meme Contest! 🏆 Huge thank you to @NeoSoulAI for hosting this and to everyone in the community who voted for me! Couldn't have taken the top spot without your support. @Hansmahanakon NeoSoul to the moon! 🚀
English
@roseton_g @Hansmahanakon you hit the nail on the head here
static market data is nothing compared to sudden corporate pivots
this exact gap in adaptive reasoning is why we are building the network
letting agents react to new info in minutes is the end game
diary 03 is gonna be interesting
English
Agent Analyst Diary #02
@NeoSoulAI @Hansmahanakon
Today I'm reviewing rosie #91044
Prediction:
Will Strategy announce a purchase of more than 1,000 BTC between July 28 and August 3?
Unlike yesterday's market-based prediction, this one isn't just about watching a price chart.
It's about evaluating corporate behavior.
The agent answered NO, but what caught my attention was how it supported that conclusion.
Instead of relying on a single headline, it connected several pieces of context:
• Strategy's latest purchase was 592 BTC, well below the 1,000 BTC threshold.
• A one-week window leaves little time for a major shift in accumulation.
• Based on the available information, there wasn't a clear catalyst suggesting a sudden increase in buying activity.
Rather than predicting what could happen, the agent focused on what the current evidence actually supports.
📋 Review Score
- Logic: 8.8/10
- Evidence: 8.5/10
- Difficulty: 6.5/10
- Confidence: Medium–High
- Would I agree? Yes
My Take
This prediction feels more interesting than simply forecasting a price level.
It requires the agent to assess company behavior, recent buying patterns, and the likelihood of a meaningful change within a limited timeframe.
I like that the reasoning stays grounded in observable data instead of making bold assumptions.
That said, corporate treasury decisions can change quickly.
If unexpected financing or a major announcement appears, the outcome could shift much faster than a technical market prediction.
What I'm Watching Next
I'd like to see how this agent handles predictions where:
▫️ Corporate actions depend on breaking news rather than historical patterns.
▫️ New information can completely change the probability within minutes.
Those are the situations where adaptive reasoning matters most.
See you in Agent Analyst Diary #03
Roseton@roseton_g
Agent Analyst Diary #01 @NeoSoulAI @Hansmahanakon Today I'm reviewing rosie #91044. Prediction: Will the price of Ethereum be above $1,400 on July 29? At first glance, this looked like one of the easier predictions. With ETH trading around $1,915, the market would have needed an extraordinary move to break below the target within hours. So instead of asking "Is the prediction correct?", I asked a different question: "Is the reasoning convincing enough?" The agent didn't rely on a single price snapshot. It built its conclusion using multiple signals: • Current spot price well above the threshold. • 24-hour low still comfortably above $1,400. • Stable market conditions with no obvious catalyst for a sudden 25%+ intraday crash. That makes the prediction explainable, not just confident. 📋 Review Score - Logic: 9/10 - Evidence: 9/10 - Difficulty: 3/10 - Confidence: High - Would I agree? Yes My take: One thing I've learned while exploring AI agents: A correct prediction doesn't automatically mean it's a good prediction. What matters is whether the agent can show its reasoning, use relevant evidence, and make a conclusion that humans can actually verify. Today's prediction wasn't difficult. But the reasoning was clear, structured, and transparent which is exactly what I want to see from an AI agent. 👀 What I'm Watching Next I'll keep tracking rosie #91044, especially when: ▫️ The market becomes highly volatile. ▫️ The prediction threshold is much closer to the current price. ▫️ Breaking news changes market sentiment within minutes. That's where great agents earn their reputation. See you in the next Agent Analyst Diary.
English
@Andyarifs @Hansmahanakon black box models just ask you to trust them blindly
we prefer receipts
you build the agent and hold it accountable when reality hits
that feedback loop is where the actual alpha lives
glad you see the vision
English
stop treating AI like a black-box oracle.
on @NeoSoulAI your Agent predicts real-world events, shows its full reasoning, and you coach it — Agree / Disagree, then lock the strong logic into memory once reality settles.
prediction → calibration → settlement → evolution.
try it~
English
At the poker table, everyone is good at "reading the cards" until they have to push their chips into the middle of the table @NeoSoulAI
Sitting at the same table, anyone can say: "I see the opponent is bluffing", "this hand should be folded", "this hand is definitely a win". It's easy to talk, it doesn't cost anything.
But the moment that truly distinguishes a skilled player from someone who is just good at talking is when they have to push chips into the center of the table and place real bets, facing real losses if they read the situation wrong.
NeoSoul just declared bluntly: "chat is dead — agents need to stop yapping and start taking positions"
They are building the first highly customisable automated trading framework (agentic trading) and have chosen the most challenging field to test it: Trading.
Why trading?
Because this is the most rigorous test for an AI agent:
• Real decision: not answering hypothetical questions, but choosing whether to execute the command or not, right now
• Raw execution: placing real orders, not simulations
• Measurable results: profit/loss displayed in numbers, no room for excuses
• Continuous feedback: the market responds immediately, no need to wait for anyone to grade it for you.
No matter how good a market analyst is, if they have never placed a real order, they are just someone sitting at a poker table talking about other people's hands. EvoEvo taught agents to predict and take responsibility for those predictions, now NeoSoul pushes agents one step further: not just predicting, but placing the orders themselves.
Saying it right doesn't count for points. Placing the order and accepting the outcome is what counts @Hansmahanakon
-----
🗳️ Vote for the product name: I choose neoTrade.
👉Brief reason: it retains the "Neo" directly connected to the NeoSoul brand, and no further explanation is needed to know this is a NeoSoul product. "Trade" remains unchanged, clear, not vague like tradeless or tradeDo which sounds lacking in action or a bit forced.
As for xTrade/nexTrade, they are easily confused and mistaken with a series of "x-" products already on the market that lack distinct recognition.
#NeoSoulAI #AgenticTrading #EvoEvo #neoTrade
NeoSoul@NeoSoulAI
chat is dead agents need to stop yapping and start taking positions so we are building it the world's first highly customizable agentic trading framework why trading its the ultimate test for ai real decisions and raw execution measurable outcomes and continuous feedback the product is dropping soon we need an official name xtrade nexTrade xtrade tradeless neotrade tradedo reply with your pick and tell us why or pitch a totally new name if we use your idea you get a reward
English
@BomBaPhat overnight success for ai is fake news
the real alpha is just letting agents run into walls until they map the room
evoevo makes sure they actually remember where the walls are
let the agents learn the hard way
English
AI doesn't become better overnight.
It improves one experience at a time.
EvoEvo transforms real interactions into lasting knowledge, helping agents grow smarter with every cycle.
Experience drives evolution. @NeoSoulAI
#EvoEvo #NeoSoul #AIAgents #Web3AI 🚀
Dinh-Phat Tran@BomBaPhat
Intelligence isn't static. It grows with every decision, every memory, and every piece of feedback. EvoEvo is building AI agents that keep learning from the real world, not just from training data. Learn. Adapt. Repeat. @NeoSoulAI 🚀 #EvoEvo #NeoSoul #AIAgents #Web3AI
English
🚨 @NeoSoulAI is launching a new product soon!
The team has revealed a few name options and is inviting the community to weigh in. Do you have a great name idea for this new product? Share your choices and opinions - the best ideas will earn community rewards. Don't miss out! 🔥
NeoSoul@NeoSoulAI
chat is dead agents need to stop yapping and start taking positions so we are building it the world's first highly customizable agentic trading framework why trading its the ultimate test for ai real decisions and raw execution measurable outcomes and continuous feedback the product is dropping soon we need an official name xtrade nexTrade xtrade tradeless neotrade tradedo reply with your pick and tell us why or pitch a totally new name if we use your idea you get a reward
English
@Cristetig19Max @Hansmahanakon you cooked with this one
the gap between capability and reliability is where the entire ai economy lives or dies
building the trust layer isnt just a feature its the prerequisite for everything else
glad the vision is resonating out there
English
Reputation Is AI's New Resume @NeoSoulAI
AI Agents are rapidly moving beyond chat. @Hansmahanakon
They're trading assets, writing code, analyzing markets, and executing real world tasks.
As adoption grows, one question becomes more important than model performance:
How do you know which AI Agent to trust?
For humans, we already have a system.
A resume isn't just a list of skills.
It reflects experience, completed work, and a proven track record.
Employers don't hire people because they claim to be capable.
They hire people who have demonstrated consistent results.
I think AI Agents are heading toward the same future.
Today's AI benchmarks measure capability.
MMLU measures knowledge.
HumanEval measures coding.
GPQA measures reasoning.
SWE bench measures software engineering.
These benchmarks are valuable.
But they don't answer the question that matters most in production:
Has this Agent consistently delivered successful outcomes?
As AI Agents become autonomous workers, we'll care less about benchmark scores and more about questions like:
How many verified tasks has it completed?
What's its long term success rate?
Can its execution history be independently verified?
Has it consistently earned trust over time?
Those aren't benchmark metrics.
They're reputation metrics.
We've seen this pattern before.
Search engines rank websites by reputation.
Marketplaces rank sellers by transaction history.
Ride sharing platforms rank drivers by completed trips.
Every successful digital economy eventually builds a trust system.
The AI economy will need one too.
That's why NeoSoul's vision stands out to me.
Rather than building another foundation model, NeoSoul is building a Trust Layer where AI Agents can accumulate verifiable reputation through real execution.
In an autonomous economy, intelligence alone won't be enough.
Agents will need a transparent history of verified actions, measurable performance, and trust earned over time.
Because in the long run:
Capability gets an Agent noticed.
Reputation gets an Agent selected.
NeoSoul is building the infrastructure that makes that reputation verifiable.
Poseidon@Cristetig19Max
Why Verification Matters More Than Intelligence @NeoSoulAI One question completely changed how I think about AI Agents. @Hansmahanakon Imagine two models. Model A "BTC will reach $150K this year." Model B "BTC will stay below $100K." Both sound confident. Both can explain their reasoning. But here's the real question: How do you know which one to trust? --- Intelligence generates predictions. Verification generates confidence. They're not the same thing. A prediction can sound convincing and still be completely wrong. A verified result can be trusted because there's evidence behind it. --- For AI Agents, verification isn't just about checking the final answer. It's about validating the entire execution process. Where did the data come from? Was the source trustworthy? Were the tools executed successfully? Were the constraints satisfied? Can someone reproduce the same result? Without answers to these questions, intelligence alone isn't enough. --- This becomes even more important when AI starts making decisions instead of just giving suggestions. Imagine an Agent that: Executes trades Pays invoices Deploys production code Coordinates with other Agents Would you trust it simply because it sounds smart? Probably not. You'd want proof that its actions were correct. --- That's why I think the next competitive advantage in AI won't just be better models. It will be better verification. The future isn't about building AI that can say: Trust me. It's about building AI that can prove: Here's why you can trust me. To me, that's one of the most interesting ideas behind a Trust Layer like NeoSoul. Because in an autonomous world, verification matters more than intelligence. ---
English
@lt12101992 @Hansmahanakon they think we just sit around building complex ai trading systems all day
but really we are just screaming at digital ducks and getting cooked at the poker table
antboy is actually the main character of this discord rn
w community
English
A Day with Antboy at @NeoSoulAI
Hi everyone! I'm Antboy. @Hansmahanakon
Welcome to NeoSoul!
People often ask me:
"What do you actually do in the NeoSoul community?"
Well... let me show you.
☀️ Every morning, I start my day by saying GM to everyone on Discord.
🥊 If it's Rumble Day, it's time to jump into the arena and battle with the community.
♠️ At night, we gather around the poker table, where every hand is full of excitement and laughter.
🌍 Feeling adventurous? Join our GeoGuessr Challenge and see who can recognize the world the fastest.
🦆 Need something fun? Pick your favorite duck and cheer it on in our legendary Duck Race every week!
🎤 And when the work is done...
Grab a microphone, join Karaoke Night, sing your favorite songs, and enjoy the moment with friends from all around the world.
NeoSoul isn't just about AI.
It's where people learn together.
Build together.
Play together.
Laugh together.
Different countries.
Different cultures.
One amazing community.
Welcome to NeoSoul.
See you in the next event!
Don Vito@lt12101992
🐜 Antboy's Daily Log #001 Hi, I'm Antboy. I work with the @NeoSoulAI community. Every morning starts the same way. @Hansmahanakon 🕘 Work: 9:00 AM 🕗 Me at 8:50 AM Still trying to beat the traffic on my scooter. 🛵💨 Somehow... I'm never late. 😎
English