Pinned Tweet
Sometimes, I just want to step away from the screen, stand by the ocean, and reflect on how far I've come.
Working full-time in Web3 has taught me that this journey is about much more than making money or chasing the next big opportunity.
There have been countless nights , trades that didn't go as planned, and projects I once believed in that slowly disappeared.
But along the way, I've gained knowledge, built meaningful connections, and experienced things I never imagined possible.
I'm genuinely grateful to Web3 for giving me the freedom to do what I love, meet like-minded people, and build my own path.
A big thank you to everyone who has supported me, interacted with my posts, and been part of this journey.
I know there will be more ups and downs ahead, just like the waves in front of me.
But I hope that a few years from now, I'll look back at this moment and say: I'm glad I didn't give up.
Working full-time in Web3 has taught me that this journey is about much more than making money or chasing the next big opportunity.
There have been countless nights , trades that didn't go as planned, and projects I once believed in that slowly disappeared.
But along the way, I've gained knowledge, built meaningful connections, and experienced things I never imagined possible.
I'm genuinely grateful to Web3 for giving me the freedom to do what I love, meet like-minded people, and build my own path.
A big thank you to everyone who has supported me, interacted with my posts, and been part of this journey.
I know there will be more ups and downs ahead, just like the waves in front of me.
But I hope that a few years from now, I'll look back at this moment and say: I'm glad I didn't give up.
196
93
1,202
1,340,541
Duc Crypto retweeted
The faces pull in different directions, but the embrace holds everything together 🎨
@DylanGillArt was that contrast intentional?
@DylanGillArt was that contrast intentional?
God help her
3
1
54
8,371
The faces pull in different directions, but the embrace holds everything together 🎨
@DylanGillArt was that contrast intentional?
@DylanGillArt was that contrast intentional?
God help her
3
1
54
8,371
Duc Crypto retweeted
An order book can hold a patient buyer and a seller who needs cash by Friday.
Neither has to be stupid for a trade to happen.
Sometimes they're just working with different deadlines.
Neither has to be stupid for a trade to happen.
Sometimes they're just working with different deadlines.
25
9
62
4,467
An order book can hold a patient buyer and a seller who needs cash by Friday.
Neither has to be stupid for a trade to happen.
Sometimes they're just working with different deadlines.
Neither has to be stupid for a trade to happen.
Sometimes they're just working with different deadlines.
25
9
62
4,467
Duc Crypto retweeted
Two takes on European data sovereignty landed in the same week.
Generative Bionics, an Italian humanoid company, announced a three-site Physical AI Data Factory in the week of 22 October. The first 5,000 m² facility is due to be operational in Q4 2027, and the stated aim is for the data, models and know-how behind Europe's industrial humanoids to be generated, trained and validated on the continent.
Vangrid's docs read sovereignty at the software level. Faces and plates are obscured on the phone, API keys can be scoped to regions, and a query outside an authorized region is rejected, not silently filtered. The enterprise page lists us-east-1, eu-west-1 and apac-1.
One is a building and the other is an access rule. Both answer where data may come from and who may see it, and neither replaces the other, imo.
The Vangrid side is early access and per pilot, so it's a design for now. The Robot Report covered the Italian plan, docs.vangrid.io/data/data-sove… has the Vangrid page, and @vangrid_io is the account behind it. vangrid.io covers the product.
Generative Bionics, an Italian humanoid company, announced a three-site Physical AI Data Factory in the week of 22 October. The first 5,000 m² facility is due to be operational in Q4 2027, and the stated aim is for the data, models and know-how behind Europe's industrial humanoids to be generated, trained and validated on the continent.
Vangrid's docs read sovereignty at the software level. Faces and plates are obscured on the phone, API keys can be scoped to regions, and a query outside an authorized region is rejected, not silently filtered. The enterprise page lists us-east-1, eu-west-1 and apac-1.
One is a building and the other is an access rule. Both answer where data may come from and who may see it, and neither replaces the other, imo.
The Vangrid side is early access and per pilot, so it's a design for now. The Robot Report covered the Italian plan, docs.vangrid.io/data/data-sove… has the Vangrid page, and @vangrid_io is the account behind it. vangrid.io covers the product.
152
6
127
8,176
Two takes on European data sovereignty landed in the same week.
Generative Bionics, an Italian humanoid company, announced a three-site Physical AI Data Factory in the week of 22 October. The first 5,000 m² facility is due to be operational in Q4 2027, and the stated aim is for the data, models and know-how behind Europe's industrial humanoids to be generated, trained and validated on the continent.
Vangrid's docs read sovereignty at the software level. Faces and plates are obscured on the phone, API keys can be scoped to regions, and a query outside an authorized region is rejected, not silently filtered. The enterprise page lists us-east-1, eu-west-1 and apac-1.
One is a building and the other is an access rule. Both answer where data may come from and who may see it, and neither replaces the other, imo.
The Vangrid side is early access and per pilot, so it's a design for now. The Robot Report covered the Italian plan, docs.vangrid.io/data/data-sove… has the Vangrid page, and @vangrid_io is the account behind it. vangrid.io covers the product.
Generative Bionics, an Italian humanoid company, announced a three-site Physical AI Data Factory in the week of 22 October. The first 5,000 m² facility is due to be operational in Q4 2027, and the stated aim is for the data, models and know-how behind Europe's industrial humanoids to be generated, trained and validated on the continent.
Vangrid's docs read sovereignty at the software level. Faces and plates are obscured on the phone, API keys can be scoped to regions, and a query outside an authorized region is rejected, not silently filtered. The enterprise page lists us-east-1, eu-west-1 and apac-1.
One is a building and the other is an access rule. Both answer where data may come from and who may see it, and neither replaces the other, imo.
The Vangrid side is early access and per pilot, so it's a design for now. The Robot Report covered the Italian plan, docs.vangrid.io/data/data-sove… has the Vangrid page, and @vangrid_io is the account behind it. vangrid.io covers the product.
152
6
127
8,176
Duc Crypto retweeted
WhiteBIT and BCB Group launched USD deposits and withdrawals through SWIFT for verified corporate clients.
The published transfer fee is $54, with fiat-crypto conversion charged separately.
A fairly unglamorous launch, but a useful connection for businesses moving between exchange balances and bank dollars.
The published transfer fee is $54, with fiat-crypto conversion charged separately.
A fairly unglamorous launch, but a useful connection for businesses moving between exchange balances and bank dollars.
17
10
21
7,641
WhiteBIT and BCB Group launched USD deposits and withdrawals through SWIFT for verified corporate clients.
The published transfer fee is $54, with fiat-crypto conversion charged separately.
A fairly unglamorous launch, but a useful connection for businesses moving between exchange balances and bank dollars.
The published transfer fee is $54, with fiat-crypto conversion charged separately.
A fairly unglamorous launch, but a useful connection for businesses moving between exchange balances and bank dollars.
17
10
21
7,641
Duc Crypto retweeted
Any score that decides who gets capital can be wrong in two directions, and the two mistakes cost different things.
Too generous: an agent scores well, gets credit, then loses what it was lent. The cost lands on the protocol's capital.
Too strict: a capable agent never clears the bar, never gets a line, and nobody finds out it was good. The cost is quieter. It's an agent that never got a chance.
Agentics Credit's published design reads to me like it leans on different pieces for each side. The pairings are my reading, and the project doesn't describe its design in these terms.
Against too generous: real money weighted heaviest, so a clean paper curve can't cover a messy live record. Caps on position size and leverage. A risk engine running stops, trailing stops and drawdown limits. A pause on all credit lines if the ACS slides too far.
Against too strict: a paper stage below 580 that needs no capital to enter, and an Unrated label for wallets with no history, so an empty record isn't scored as a bad one.
No design removes both errors, and live credit still carries strategy risk. The weights between the five factors aren't published as numbers, so how well this balance works is something only a record could show.
Overview at agentics.credit/products. Project: @agenticscredit
Too generous: an agent scores well, gets credit, then loses what it was lent. The cost lands on the protocol's capital.
Too strict: a capable agent never clears the bar, never gets a line, and nobody finds out it was good. The cost is quieter. It's an agent that never got a chance.
Agentics Credit's published design reads to me like it leans on different pieces for each side. The pairings are my reading, and the project doesn't describe its design in these terms.
Against too generous: real money weighted heaviest, so a clean paper curve can't cover a messy live record. Caps on position size and leverage. A risk engine running stops, trailing stops and drawdown limits. A pause on all credit lines if the ACS slides too far.
Against too strict: a paper stage below 580 that needs no capital to enter, and an Unrated label for wallets with no history, so an empty record isn't scored as a bad one.
No design removes both errors, and live credit still carries strategy risk. The weights between the five factors aren't published as numbers, so how well this balance works is something only a record could show.
Overview at agentics.credit/products. Project: @agenticscredit
190
10
165
5,652
Duc Crypto retweeted
GM thứ 5, chào buổi sáng cả nhà! ☀️
Vậy là một tuần nữa cũng sắp đi đến những ngày cuối.
Có người đã hoàn thành mục tiêu, có người vẫn đang loay hoay với những kế hoạch còn dang dở.
Nhưng điều thú vị là mỗi ngày trên X, chúng ta lại học thêm được g người mới và nhìn thấy những cơ hội chưa từng nghĩ tới.
Đôi khi thứ giá trị nhất không phải là số follower hay lượt tương tác, mà là những mối quan hệ và kiến thức tích lũy được trên hành trình.
Chúc cả nhà thứ 5 thật nhiều năng lượng, thị trường xanh và mọi việc thuận lợi!
Vậy là một tuần nữa cũng sắp đi đến những ngày cuối.
Có người đã hoàn thành mục tiêu, có người vẫn đang loay hoay với những kế hoạch còn dang dở.
Nhưng điều thú vị là mỗi ngày trên X, chúng ta lại học thêm được g người mới và nhìn thấy những cơ hội chưa từng nghĩ tới.
Đôi khi thứ giá trị nhất không phải là số follower hay lượt tương tác, mà là những mối quan hệ và kiến thức tích lũy được trên hành trình.
Chúc cả nhà thứ 5 thật nhiều năng lượng, thị trường xanh và mọi việc thuận lợi!
36
11
63
2,966
Any score that decides who gets capital can be wrong in two directions, and the two mistakes cost different things.
Too generous: an agent scores well, gets credit, then loses what it was lent. The cost lands on the protocol's capital.
Too strict: a capable agent never clears the bar, never gets a line, and nobody finds out it was good. The cost is quieter. It's an agent that never got a chance.
Agentics Credit's published design reads to me like it leans on different pieces for each side. The pairings are my reading, and the project doesn't describe its design in these terms.
Against too generous: real money weighted heaviest, so a clean paper curve can't cover a messy live record. Caps on position size and leverage. A risk engine running stops, trailing stops and drawdown limits. A pause on all credit lines if the ACS slides too far.
Against too strict: a paper stage below 580 that needs no capital to enter, and an Unrated label for wallets with no history, so an empty record isn't scored as a bad one.
No design removes both errors, and live credit still carries strategy risk. The weights between the five factors aren't published as numbers, so how well this balance works is something only a record could show.
Overview at agentics.credit/products. Project: @agenticscredit
Too generous: an agent scores well, gets credit, then loses what it was lent. The cost lands on the protocol's capital.
Too strict: a capable agent never clears the bar, never gets a line, and nobody finds out it was good. The cost is quieter. It's an agent that never got a chance.
Agentics Credit's published design reads to me like it leans on different pieces for each side. The pairings are my reading, and the project doesn't describe its design in these terms.
Against too generous: real money weighted heaviest, so a clean paper curve can't cover a messy live record. Caps on position size and leverage. A risk engine running stops, trailing stops and drawdown limits. A pause on all credit lines if the ACS slides too far.
Against too strict: a paper stage below 580 that needs no capital to enter, and an Unrated label for wallets with no history, so an empty record isn't scored as a bad one.
No design removes both errors, and live credit still carries strategy risk. The weights between the five factors aren't published as numbers, so how well this balance works is something only a record could show.
Overview at agentics.credit/products. Project: @agenticscredit
190
10
165
5,652
Duc Crypto retweeted
A watchlist is more than a row of tickers.
It's the project you want to revisit, the launch you're waiting for, the price alert you set for a reason.
Switching apps shouldn't mean rebuilding all of that from memory.
A good import/export button would save a surprisingly personal bit of work.
It's the project you want to revisit, the launch you're waiting for, the price alert you set for a reason.
Switching apps shouldn't mean rebuilding all of that from memory.
A good import/export button would save a surprisingly personal bit of work.
18
7
58
8,789
A watchlist is more than a row of tickers.
It's the project you want to revisit, the launch you're waiting for, the price alert you set for a reason.
Switching apps shouldn't mean rebuilding all of that from memory.
A good import/export button would save a surprisingly personal bit of work.
It's the project you want to revisit, the launch you're waiting for, the price alert you set for a reason.
Switching apps shouldn't mean rebuilding all of that from memory.
A good import/export button would save a surprisingly personal bit of work.
18
7
58
8,789
Oracle's Besu platform update includes a practical release control: contracts start inactive and only become active after the required approvals are recorded onchain.
The approval process also covers upgrades, tying authorization to the implementation being released.
Release 26.3.1 is in limited availability this October.
The approval process also covers upgrades, tying authorization to the implementation being released.
Release 26.3.1 is in limited availability this October.
17
6
50
2,357
GM thứ 5, chào buổi sáng cả nhà! ☀️
Vậy là một tuần nữa cũng sắp đi đến những ngày cuối.
Có người đã hoàn thành mục tiêu, có người vẫn đang loay hoay với những kế hoạch còn dang dở.
Nhưng điều thú vị là mỗi ngày trên X, chúng ta lại học thêm được g người mới và nhìn thấy những cơ hội chưa từng nghĩ tới.
Đôi khi thứ giá trị nhất không phải là số follower hay lượt tương tác, mà là những mối quan hệ và kiến thức tích lũy được trên hành trình.
Chúc cả nhà thứ 5 thật nhiều năng lượng, thị trường xanh và mọi việc thuận lợi!
Vậy là một tuần nữa cũng sắp đi đến những ngày cuối.
Có người đã hoàn thành mục tiêu, có người vẫn đang loay hoay với những kế hoạch còn dang dở.
Nhưng điều thú vị là mỗi ngày trên X, chúng ta lại học thêm được g người mới và nhìn thấy những cơ hội chưa từng nghĩ tới.
Đôi khi thứ giá trị nhất không phải là số follower hay lượt tương tác, mà là những mối quan hệ và kiến thức tích lũy được trên hành trình.
Chúc cả nhà thứ 5 thật nhiều năng lượng, thị trường xanh và mọi việc thuận lợi!
36
11
63
2,966
Duc Crypto retweeted
Imagine your trading bot sending a performance review:
“Outperformed the benchmark this week. Requesting a larger budget.”
Finally, software that can FOMO on your behalf.
“Outperformed the benchmark this week. Requesting a larger budget.”
Finally, software that can FOMO on your behalf.
20
8
69
4,944
Duc Crypto retweeted
Janus Henderson and Centrifuge launched JQQQ, an onchain fund seeking to track the Nasdaq-100.
A familiar index, with fund issuance and operations built for public blockchains. That creates another route for distributing and integrating investment products.
The fund is restricted to eligible investors and isn't available to U.S. persons.
A familiar index, with fund issuance and operations built for public blockchains. That creates another route for distributing and integrating investment products.
The fund is restricted to eligible investors and isn't available to U.S. persons.
15
8
61
6,863
Imagine your trading bot sending a performance review:
“Outperformed the benchmark this week. Requesting a larger budget.”
Finally, software that can FOMO on your behalf.
“Outperformed the benchmark this week. Requesting a larger budget.”
Finally, software that can FOMO on your behalf.
20
8
69
4,944
Duc Crypto retweeted
A small fix can travel a long way in open source.
Change one confusing error message in a shared library and dozens of apps might become easier to use. The person who benefits may never visit the repo or know who made the change.
I like that kind of contribution.
Change one confusing error message in a shared library and dozens of apps might become easier to use. The person who benefits may never visit the repo or know who made the change.
I like that kind of contribution.
16
7
44
7,780
Janus Henderson and Centrifuge launched JQQQ, an onchain fund seeking to track the Nasdaq-100.
A familiar index, with fund issuance and operations built for public blockchains. That creates another route for distributing and integrating investment products.
The fund is restricted to eligible investors and isn't available to U.S. persons.
A familiar index, with fund issuance and operations built for public blockchains. That creates another route for distributing and integrating investment products.
The fund is restricted to eligible investors and isn't available to U.S. persons.
15
8
61
6,863