A decentralized AI compute play just outran 99% of the crypto market this month
$AKT, the native token of @akashnet, is up 72% on the month and leading CoinMarketCap's Broad Market category despite a 12% pullback Saturday testing $0.595 support. The DePIN GPU marketplace lets users rent NVIDIA H100s for $1.49/hour vs $43K outright.
Founder Greg Osuri presented at the Economist Impact AI Compute Summit in Amsterdam this week. Recent integrations with Venice ai and FLock io target decentralized AI inference, the same vertical NEAR and Sahara AI are also rotating into.
Potential bottleneck with a powerful advantage in terms of moat and licensing:
$LEU (Centrus Energy) — ~$5.8B market cap
They are the ONLY company in the United States currently licensed to enrich uranium to HALEU levels.
Let that sink in.
HALEU = High-Assay Low-Enriched Uranium. The fuel that powers next-generation SMR (Small Modular Reactor) nuclear plants.
Every advanced reactor company building for the AI power boom needs it.
There is exactly one domestic supplier.
Here’s why this is a bigger deal than people realize:
- The DOE just announced new efforts to boost the nuclear fuel supply chain
- SMRs are the leading long-term solution to the AI data center power crisis
- No HALEU = no SMRs. It’s that simple.
- There is no near-term domestic alternative. The supply chain runs through $LEU or it doesn’t run.
The bottleneck thesis in one sentence:
You can build all the next-gen reactors you want. Without HALEU, they’re just expensive concrete.
$LEU is the pick-and-shovel play hiding in plain sight at the base of the entire U.S. nuclear renaissance.
HFT algorithms are already reading posts about this ticker and moving it 8%+ on DOE news days.
The institutions will figure it out eventually.
The question is whether you do first.
-----
*Not financial advice. Do your own due diligence.
$LEU has $2.3 billion in commercial backlog and a $900 million DOE HALEU award.
It holds $2 billion in cash. It is the only US-origin enrichment technology that is deployment-ready for both commercial and national security programmes.
Congress banned the Russian alternative.
The market is still pricing this as a broker-trader business. It is becoming a domestic enrichment monopoly.
One more detail buried in the shareholder letter: Palantir is embedded in the build-out. Early work has already identified $300 million in potential cost savings.
The chips got repriced. The power got repriced.
The fuel still hasn't. $LEU $PLTR $CCJ $UEC
Donald Trump attacks Giorgia Meloni — and she delivers a fiery speech he’ll never forget.
Donald Trump thought he could easily score political points by calling Giorgia Meloni “an insult to Jesus,” accusing her of “not being woke,” and claiming that God does not discriminate. Unfortunately for “Don Dementia,” this time he picked the wrong target.
Standing at a historic location, Giorgia Meloni didn’t just respond — she delivered a full moral reckoning.
“The President of the United States just said that I insult Jesus,” Giorgia Meloni declared. “Do you want to know what really insults Jesus? Taking healthcare away from the sick while cutting taxes for billionaires.”
And that was only the beginning.
“Do you want to know what else insults Jesus?” she continued. “Deporting foreigners and separating children from their mothers.”
Then she went even further, touching on war, corruption, and hypocrisy.
“Do you want to know what insults Jesus? Bombing innocent schools in Iran and sending our brave men and women to die in yet another endless war… hiding the Epstein files and then refusing to prosecute anyone involved.”
This wasn’t politics as usual. It was a full moral indictment.
Giorgia Meloni, targeted by Trump for supporting transgender people and for saying that “trans children are children of God,” completely turned the tables. Instead of backing down, she grounded her message in the very teachings Trump tried to weaponize.
“I am not a perfect Christian,” she said. “There was only one perfect man, and two thousand years ago he was crucified.”
Then came the line that hit the hardest:
“Jesus told us to love our neighbor as ourselves… Can we imagine war in heaven? Can we imagine hatred in heaven? Can we imagine poverty in heaven? Then why do we tolerate these things on Earth?”
This is how you respond. Not with insults. Not with fear. But with clarity and conviction.
Trump tried to discredit her. Instead, Giorgia Meloni delivered a sermon that now echoes far beyond that hall.
Please share Giorgia Meloni’s inspiring words.
@gregosuri yes, I use it every day, and pay almost$ 10 daily to them, sometimes more, ever single day. Also on weekends the free and cheap models always get a API rate limit, I think I am not the only one using it 😄. Why not paying to Akash?
1/ Deploying AI agents today typically requires navigating complex cloud configurations, managing GPU provisioning, and dealing with opaque pricing before writing a single line of agent logic.
A developer who simply wants to test an agent framework like Hermes or OpenClaw must first set up infrastructure that has nothing to do with their actual work. The friction is enormous, and the margin for error is high.
@akashnet ....and still it needs to be used, somehow the customers do not find the easy way to setup and wait 3-4 months for a mac studio which cannot be adjusted along the way. how can you improve your marketing?!
The Mac mini was never the move. The OpenClaw template on Akash Console runs on decentralized compute for a fraction of the cost, and Akash Agents (https://t.co/PUMstm0iQj) simplifies the setup process even further.
The $800 hardware was just a distraction.
Greg Osuri built the case for decentralized AI in three layers in under 5 minutes.
Layer one: access.
Anthropic's Claude is the best coding model available. OpenClaw is the most successful open-source project in history. Anthropic just announced they're banning OpenClaw usage. The most powerful model cutting off the most loved tool. The company that controls the best intelligence can revoke your access whenever their interests change.
Layer two: transparency.
You have no idea what biases Anthropic has built into Claude. If AI models directly impact how you think, the decision tree needs to be visible. Closed-source models with hidden biases making decisions that shape your worldview isn't a feature, it's a risk.
Layer three: compute.
Amazon may be picking winners in AI by controlling compute allocation. Cloud companies stopped being neutral marketplaces and started giving GPUs to the companies they've bet on. Amazon overwhelmingly backing Anthropic might have contributed to Claude's dominance. The cloud is no longer a utility. It's a kingmaker.
@gregosuri's conclusion: intelligence should be free like water and air.
Decentralized AI isn't ideology. It's the only architecture where the model can't ban your tools, the biases can be audited, and the compute isn't allocated by a company picking favorites.
@RHerman@Gregson36673847 no worries, I found one , which is truly open source, will focus on something else, that is truly open source and not close source like yours https://t.co/DRF3OMBD0A
🚨 IT’S LIVE 🚨
After weeks of building, testing, and working directly with @TradingView moderators…
Multi Timeframe Fractal Map [Herman] is officially RELEASED FREE. 🎉
🔥 BONUS:
If this post hits 50 retweets I’ll release the FULL script as OPEN SOURCE
This is the tool traders have been asking me about.
And now you have it.
What it gives you:
• higher timeframe context (5m → Daily)
• real-time liquidity sweeps
• SMT across correlated markets
• execution structure on lower timeframes
• Multi timeframe CISD
•PSP candle logic
All on ONE chart.
This wasn’t rushed.
It went through direct review with TradingView moderators and is fully compliant.
This is just the beginning.
Question:
If you had this as open source…
what would you add, change, or build on top of it? 👇
#NQ #FuturesTrading #ICT
👇 Access → link in first comment
@vulturesxbt ich könnte dir sogar deine trade challenge backtesten auf 20-30 pairs , um überhaupt zu wissen, ob es die Mühe sich lohnt, falls ich da Zugriff bekomme
@RHerman@Gregson36673847 OK, ....and where is the link to the open source code? I am asking because like this I can backtest it and see if it is worth investing time in it. I can share my backtests over multiple pairs and timeframes. This is how open source works, multiple people work on it🙂