Early-stage Norwegian web3 startup, building personal agents and radically simple DeFi interfaces. Telegram: https://xitter.cf/t.co/iI5wsM0XVJ
By Solberg Invest
Joined May 2025
- Tweets1.7K
- Following17
- Followers1.2K
- Likes163
Pinned Tweet
The past 6 months, we've been building.
Now we're gradually shifting our focusing to growth.
We've got 100+ personal agents running.
Goal is 1,000 paying users in 2026.
Revenue from agents will fuel a company treasury of $BMIND tokens.
When trading goes live (pending MiCA license approval), 20% of trading fees will be used to buy and burn $BMIND
We're hosting monthly live calls, sharing our progress.
If you want invitations to these calls, join the telegram group - link in the bio.
On the product end, we're going to incorporate technologies developed by Bittensor subnets (like @SomaSubnet and @hippius_subnet), and we're exploring @AskVenice to enhance privacy for users.
- We're part of K2, a Norwegian incubator.
- We're in the Google Cloud Startup Program.
- We're the first-ever Web3 company to get a $100K grant from Innovation Norway.
- We've got @SolbergInvest with +10K newsletter subscribers rooting for us.
- We've got some of the most gifted senior devs in Norway on the team.
... we're getting a head start on the next bull.
P.S.
Docs etc. are outdated.
Subnets are goldmines for cheap, permissionless quality tech.
A Bittensor subnet just built an AI safety model that beats the big players
Trishool (@trishoolai), Bittensor's (@opentensor) subnet 23, released HaloGuard 1.0 today. The 4B version claims first place across seven established prompt-safety benchmarks, and the 0.8B version outperforms models several times its size.
The pitch is speed over scale. HaloGuard screens prompts in real time before they reach the underlying model or agent, instead of catching harmful outputs after the fact.
An earlier alpha version already runs live on the Chutes subnet, hitting an 87% F1 score on real traffic since May.
The subnet's whole design is built around attacking itself. Miners get paid for breaking the model, then the breaks get patched.
BlockMind retweeted
Introducing The Grid Atlas, an interactive globe that visualizes where Web3's rails actually reach.
Explore the map at map.thegrid.id and see Web3's geography for yourself. Our new product country support data now powers the Atlas. Log in and add your country support to help make the map even more complete.
The ETF bid is visible.
The confirmation is not.
BTC has a loud spot ETF narrative again, but today’s venue data is more cautious:
Coinbase premium flipped negative.
Funding is negative almost everywhere.
OI is shrinking.
That gap matters. 1/5
Aave paused buybacks after the rsETH incident.
That is not a small signal. It means DeFi’s trust problem has reached the treasury/governance layer.
But the capital data says something more nuanced. 1/5
This is the distinction that matters:
DeFi risk is not only price risk. It is dependency risk.
When a restaking asset gets interwoven enough that a DAO needs treasury flexibility, the APY was never the full story. 4/5
That's the part I think matters.
In DeFi stress events, 'safe' is starting to mean more than clean reserves. It also means who can freeze, backstop, and restore liquidity when something breaks.
Different hierarchy, different winner. 4/5