Tech guy with a blueprint and a pulse. Building @firsttensor
Europe
Joined December 2023
- Tweets347
- Following397
- Followers894
- Likes2.1K
Pinned Tweet
Back in 2022 I've invested a shit amount of money and time into $TAO as I saw the potential right there (thanks Dread).
I've invested my own savings and also I was crazy enough to get the biggest loan that I could from ING Romania and I've invested ALL of it into #Bittensor.
In 2023 I've decided to quit my DevOps Engineer job and get back my time and freedom. I've managed to do all that with the help of TAO.
Fast forward almost 5 years, it was all worth it!
$TAO above $300, 1st halving done, almost 4 billion marketcap, 128 subnets, and revenue increasing each month.
Now in 2026 I've invested a shit amount of money and I plan to invest a shit amount of time also into #GenesisL1!
Why? Because I can see the potential right here! I can see the roadmap, I can see what the project can do in the next 2-3 years. It gives me the TAO feelings!
At $5mil marketcap is the next big thing right there!
You might want to learn more about it and invest if you also see the potential in it.
Cheers!
Eseck retweeted
The AI world is moving into the right direction and will help hundreds if not billions of people in the coming years directly or indirectly.
That’s why #GenesisL1 is so important into this space! genesisl1:native
this might be my last normal tweet before biotech goes vertical
i'm not a whistleblower. I don't work at Lilly, Novo or any of the big labs
but there is a massive shift happening in human biology right now.
it's sitting in press releases and almost nobody reads them
the last 6 months alone:
- a once-daily pill roughly doubled median survival in metastatic pancreatic cancer (13.2 vs 6.6 months)
- a one-time gene editing infusion cut hereditary angioedema attacks 87%, and 62% of patients stayed attack-free and off all other prevention for 6 months
- a one-time gene edit for cholesterol dropped LDL up to 62%, and it held for over a year
- a personalized mrna vaccine beat keytruda alone in a 1,137 patient melanoma phase 3
- retatrutide hit ~28% weight loss in a phase 3
- an ai-designed drug entered phase 3
- a pill that regrows hair passed its second phase 3
- an immunotherapy that could let some rectal cancer patients skip surgery, chemo and radiation just got fda approved
- a man lived 271 days off dialysis on a pig kidney
- human cells were caught passing chromosome-sized chunks of DNA to their neighbors, and it changes how the neighbors behave
we are on the verge of an era where the human body is a software problem and the update cycle gets shorter everyday
stick around, it's going to get weird
bio/acc
Eseck retweeted
This is HUGE.
And this is exactly why I keep telling people to pay attention to @genesis_L1.
Imagine AI agents having access to over 100 MILLION biological sequences, the entire human reference genome, and tools to search protein structures, compare 3D folds, analyze sequences and much more.
Well… that’s the scale GenesisL1 is aiming for.
Its blockchain explorer, GL1X, already gives AI agents access to over 1 MILLION molecular records and 43 bioinformatics tools.
Also planning to add 100 MILLION+ sequences and the entire human genome on-chain.
Let that sink in. genesisl1:native
“If we can predict the [protein's] shape, we can design a drug that recognizes the harmful form and leaves the healthy one alone.”
A Hopkins-led team backed by up to $52M from @ARPA_H will build an AI model to decode how proteins misfold in Parkinson’s, cancer, and other diseases. bit.ly/4zMyCdb
Eseck retweeted
Feels like a new narrative is starting to take shape.
Science!
AI-driven drug discovery is going to be HUGE.
As AI agents become increasingly involved in scientific research, they'll need more than intelligence.
They'll need access to scientific data, computational tools, reproducible results and verifiable provenance.
This is where @genesis_L1 gets very interesting.
Its GL1X explorer gives AI agents access to over 1 MILLION molecular records anchored on GenesisL1 and 43 bioinformatics tools.
Protein structures. Sequence analysis. Molecular comparisons. Scientific literature.
This could eventually become one of the most powerful narratives in crypto.
Drug discovery. Biotechnology. Longevity. Open science.
Think about the amount of value that could be created in these industries.
Learn GenesisL1. genesisl1:native
Don't be lazy.
Eseck retweeted
This blockchain explorer is absolutely MIND-BLOWING. 🤯
Seriously. Take a look at the GL1X explorer. genesisl1:native
This is not your typical blockchain explorer where you check transactions, wallets and validators.
This is a blockchain explorer combined with a working BIOINFORMATICS LAB.
Let that sink in.
Over 1 MILLION molecular records living on-chain.
Protein structures. Molecular data. AI models. Scientific research.
And here’s where things get CRAZY.
Connect Claude or ChatGPT through MCP and your AI agent gets access to 87 tools, including 43 for bioinformatics.
Your agent can explore molecular structures, compare proteins, analyze peptides, search scientific literature and build datasets for AI models.
All through a blockchain explorer.
No API key. No sign-up.
And the entire thing is OPEN SOURCE.
WTF???
Now think about something.
If this is what a BLOCKCHAIN EXPLORER looks like on GenesisL1…
Can you even imagine what the entire infrastructure could become?
And that’s exactly what I’ve been trying to explain for months.
GenesisL1 is not just another blockchain trying to find a use case.
It’s an infrastructure being built around science, AI, verifiable data, computation, intellectual property and machine payments.
This explorer alone gives you an ideea of how much bigger this vision is than the typical crypto narrative.
I believe most people have absolutely no idea what they’re looking at when they first discover GenesisL1.
Eseck retweeted
We are upgrading the genesisl1:native onchain #Biotech insilico laboratory a few versions every week with amazing new features and improvements gl1x.com/molnft. Today, GenesisL1 stores >1M molecular records. A year from now: >100M sequences & the entire Human Genome onchain. #DeSci
Eseck retweeted
Imagine giving an AI agent access to an entire blockchain AND a scientific laboratory.
Now look at GL1X built on genesisl1:native
1M+ molecular records. Bioinformatics. On-chain AI. 87 tools for AI agents.
#GenesisL1 isn't just putting AI on-chain. It's making the chain usable by AI.
So, instead of having 10-15 people in the jury, you now have only ONE that can decide the fate of someone (miner rewards), niiiice.. bittensor:native
POW registrations and +2500 miners per subnet could have been achieved without NULL consensus.
This just opens the door for subnet owners that also runs a validator and miners to just do what they want to do there.
It’s like you’re the judge, the jury, and the executioner, while also being the defendant.
What could possibly go wrong?
Bittensor = Centralizensor
Eseck retweeted
How to launch your first crypto AI model on GenesisL1.
NO CODING REQUIRED! 🤯
Here’s a quick guide to GL1F Crypto 👇
1️⃣ CHOOSE YOUR DATA
Open crypto.gl1f.com → Create Crypto AI model.
Select:
• Exchange: where the data comes from.
• Market: the coin you want to study.
• Candle: data resolution - 15m means one candle every 15 minutes.
• Start / End: the historical period your model will study.
For this example: Binance → ETH → 15m candles → six months of completed history.
2️⃣ DEFINE YOUR QUESTION
Use these example settings:
• Direction: Up
• EMA baseline: 5
• Target: 1%
• Stop: 0.5%
• Horizon: 5 hours
EMA is a moving average that gives more weight to recent prices. “5” means five candles.
Your question becomes:
“Will ETH reach 1% above EMA5 before reaching 0.5% below it, within the next five hours?”
Both levels are measured from the EMA baseline.
The model learns from historical yes/no outcomes.
3️⃣ CHOOSE YOUR SIGNALS
Signals are the information the model can see: price changes, volatility, volume, momentum, BTC movements and more.
Start with the “Optimal” preset.
More signals don’t automatically mean a better model.
Click “Build dataset” → “Continue to training”.
4️⃣ TRAIN YOUR MODEL
For your first experiment, keep the default model settings and use “Walk-forward” validation.
That means learning from earlier data and testing on what happened next.
The main settings:
• Trees: how many learning steps the model can take.
• Depth: how complex each tree can become.
• Learning rate: how much each tree contributes.
• Min leaf: minimum examples supporting a final tree decision.
• Final test %: the latest data reserved for evaluation.
Keep early stopping enabled: it stops adding trees when validation stops improving.
Train, inspect the results, then save your .gl1f file.
5️⃣ BACKTEST WITH COSTS
Use the final test period and include fees and slippage.
“Entry threshold” is the minimum probability required for a simulated signal.
0.60 = at least 60%.
Start with 1× leverage for this exercise.
Check drawdown, trade count and the full equity curve - not just the final return.
Avoid repeatedly changing settings to make the same final test look better.
6️⃣ RUN INFERENCE
Inference means asking your trained model to evaluate a new completed candle.
A result of 0.72 means:
“The model estimates a 72% probability of the specific event you defined.”
It is an estimate, not a guarantee.
7️⃣ OPTIONAL: PUBLISH
When deployment is available, you can mint your model as an NFT on GenesisL1, choose its license and configure free access, tips or paid usage.
Pick a market. Define a question. Build. Test. Learn.
No code required. Fees are burnable.
Experimental research software. Backtests are simulations; the platform explicitly excludes live trading and automated execution. 🤝
genesisl1:native
Does #GenesisL1 store huge AI models directly on the blockchain? genesisl1:native
No, and it doesn't need to.
Large model weights and artifacts can be stored and replicated through IPFS.
GenesisL1 only needs the small references and metadata required to identify, tokenize and verify those models.
The heavy computation also doesn't happen on-chain.
It runs on the GPUs of independent operators.
So the blockchain doesn't need to store 100GB models or generate every AI token.
It coordinates the models, operators, payments, reputation and verification around them.
That's a much more scalable way to bring large AI models into Web3.
🤖 Made with AI
Eseck retweeted
Could anyone with a GPU become an AI inference provider on GenesisL1? genesisl1:native
That's the idea.
Imagine you have an RTX 3090.
You could pull a tokenized Qwen model from the GenesisL1 ecosystem, verify its artifacts, load it locally and offer inference through the marketplace.
Users could compare you with other operators based on price, uptime, performance and reputation.
User pays genesisl1:native → GPU runs inference → user gets the result.
Instead of GPUs sitting behind a centralized AI provider, GenesisL1 can connect independent compute into an open AI marketplace.
🤖 Made with AI
You are not ready for what is coming to the GenesisL1 ecosystem in the coming days.
You are not ready for what is coming in the coming months.
genesisl1:native
The next billion-dollar discovery won't be made by a human. genesisl1:native
It's gonna be made by an AI agent! 🧬
Why could genesisl1:native hit $5,000+ per token?
I know. That’s roughly a 20,000x from the current price.
Sounds insane.
But that’s “only” around a $200B market cap.
Now think about what Genesis is actually being built for.
Let’s talk about science first.
A single successful drug can create an enormous amount of economic value.
One successful peptide could be worth more than entire crypto ecosystems.
A single biomedical breakthrough can create billions in value, and GenesisL1 is being built as infrastructure for exactly these kinds of real-world scientific and computational workflows.
Models. Molecular structures. Scientific IP. Verifiable computation. Reproducible research.
Now let’s go further.
We have GL1F - Pumpfun for machine learning.
Pumpfun removed the coding barrier for launching tokens.
GL1F is removing the coding barrier for building machine-learning models.
This can go crazy. Could bring an entirely different type of user on-chain.
Then we have a decentralized Hugging Face + inference market.
Genesis is also building infrastructure where mainstream AI models can become tokenized Web3 objects and independent operators can provide inference.
Users and agents discover models → choose an operator → request inference → pay natively.
With x402, this becomes even more interesting:
machines paying machines for intelligence.
Machine-to-machine payments and inference could create an entirely new category of network activity.
Then add AI agents.
Humans might run 10 experiments.
Agents can run thousands.
Discover a dataset. Test a hypothesis. Train a model. Call another model. Pay for inference. Analyze the result. Repeat.
UI for humans. API for agents.
And all of this connects back to $L1.
And with the planned GL1F fee model, $L1 collected through those fees is intended to be automatically burned.
So now zoom out.
Machine learning + AI inference + agents + science + biotech + tokenized IP + machine payments.
Real-world utility.
Real demand.
Not just crypto speculation.
Learn GenesisL1. 🤝
Eseck retweeted
Pumpfun for machine learning. ⏳
It’s being built now, and it will burn genesisl1:native tokens.
I know this is HUGE, but at the same time, it’s just a little piece of what can be built on the GenesisL1 infrastructure.
Learn GenesisL1.
Smart money = accumulating genesisl1:native
Nearly $800 million in total cumulative revenue since its launch.
That’s Pumpfun.
They broke the coding barrier for launching tokens.
We know 99.9% of them are rugs lol.
Now imagine breaking the coding barrier for machine learning. 🤯
Anyone being able to launch their own AI model with no skills required and being able to tokenize & monetize it if they want to.
Same principle. Completely different outputs.
Well, I know this sounds fascinating, but I’m sure that, at first glance, you can’t imagine how huge this could become.
So let me tell you this…
Keep your eyes on genesisl1:native
Mikhail is cooking.
