Sabitlenmiş Tweet
I just published Idle USDC Query on Dune Analytics link.medium.com/I0GVV0NZtJb
English
giancarloangulo IBRL 🇵🇭☁️🔥💃
33.3K posts
giancarloangulo IBRL 🇵🇭☁️🔥💃
@giancarloangulo
Too old for this stuff Medium: https://t.co/fEOemVVeao
Republic of the Philippines Katılım Aralık 2021
3.5K Takip Edilen787 Takipçiler
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
For over a decade, we’ve accepted that end-to-end backprop is the only way to train deep networks. But holding the entire network in memory all at once is why AI training is hitting a resource wall.
We found a new way to break the network into blocks and train them independently. The trick? Treating the network’s forward pass like a diffusion model denoising a signal.
This reinterpretation slashes the memory needed to train deep models. In our #ICLR2026 paper (arxiv.org/abs/2506.14202), we matched end-to-end performance across ViTs, DiTs, and LLMs. We did this while training just one isolated block at a time.
Sakana AI@SakanaAILabs
Introducing DiffusionBlocks: Block-wise Neural Network Training via Diffusion Interpretation pub.sakana.ai/diffusionblocks What if we didn’t have to hold an entire neural network in memory to train it? Standard neural net training optimizes all parameters jointly. As a result, the memory required during training grows linearly with the depth of the network. In our #ICLR2026 paper, we propose DiffusionBlocks, a principled framework to train networks one block at a time, drastically reducing memory requirements while matching end-to-end performance. With DiffusionBlocks, we split the network into blocks and train them one at a time, so you only need memory for a single block. How? We explicitly assign each block a role: to move the representation a little closer to the target than the block before it did. That role turns out to be precisely what a diffusion model does, step by step. Each block only needs to optimize its own objective and can be trained independently. We validated this across five different architectures: • ViT • DiT • Masked diffusion • Autoregressive transformers • Recurrent-depth transformers In each case, performance is competitive with end-to-end training while using a fraction of the memory. This perspective also extends naturally to recurrent-depth (Looped) transformers, which apply the same network iteratively and normally require expensive backpropagation through time (BPTT). Viewed through DiffusionBlocks, we can replace those multiple iterations with a single forward pass during training. Read our paper and code, to learn more. Paper: arxiv.org/abs/2506.14202 GitHub: github.com/SakanaAI/Diffu… 🐟
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
Penang's rise into a semiconductor superpower was almost accidental. Every Penangite knows the story — few outside the island do, even Malaysian.
The Malaysian government cancelled Penang's free port status in 1969, unemployment spiked to 20%. In a moment of desperation, the state went all-in on semiconductors and built a Free Trade Zone.
Intel opened its first-ever offshore plant there in 1972. Restaurant workers were retrained for the assembly line.
Quietly, across the 80s and 90s, it became a low-key powerhouse. Fly into Penang today and you feel it — the airport is underfunded and crowded, but drive into town and you pass the who's who of semicon, manufacturing and biotech. Intel, Micron, Agilent Omicron etc. Penang runs so deep in chips that even Broadcom's Hock Tan is a Penang boy.
One caution on Malaysian semis: most Penang-listed names are OSAT assembly and test. Record YoY revenue, but compressed margins and no real moat. Be careful.
And Penang is never complete without its legendary Char Kway Teow.
Eat spicy kway teow, trade hot stocks!
NoLimit@NoLimitGains
Everyone is hunting for the next semiconductor winner. But Penang, Malaysia quietly packages 23% of every American-bound chip and runs 15% of the global OSAT and test market. Three names sit at the center of it. Here they are:
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
OFW remittances growth declines for a 3rd straight month. Expect April to be significantly worse than March, as the global energy crisis sinks its teeth deeper into household spending worldwide. Peso likely to weaken further in May on eroding investor confidence and deteriorating balance of payments. $EPHE
bloomberg.com/news/articles/…
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
My honest thoughts about the Solana ecosystem right now:
Even though I’m still extremely bullish, the ecosystem genuinely feels very different from before.
We’ve built more apps, improved infra, onboarded more users, more institutions, more capital, and generated more revenue than ever before.
The ecosystem is objectively stronger.
But somehow it also feels more lonely than before.
The space got so big that many of us don’t even know each other anymore. Things feel more artificial, distant, and transactional.
Back in 2021, the community itself carried Solana on its back. The devs were eating glass 24/7, creators onboarded users out of pure conviction, and the narratives were built organically by the people.
The foundation barely even had a social presence back then. The community had the power.
Now it feels like too many people rely on the foundation, engagement algorithms, institutional validation, and coordinated narratives to shape the culture for them. I mean "Solana kols" are now foundation members.
And that’s where I think things become dangerous.
If the narratives of the ecosystem are artificially constructed and amplified from the top down, the culture slowly loses authenticity. People stop thinking for themselves and just start echoing whatever is placed in front of them.
The culture became bigger, but weaker. It's more efficient forsure, but less human.
A lot of people no longer love Solana itself. They just use Solana because the tech is good or because the current meta is here.
There’s less collective identity, less emotional attachment to the chain, and less genuine support between builders unless there’s financial incentive attached to it.
Back then we all felt responsible for pushing Solana forward together.
Now many people just spectate lol.
The Solana I fell in love with was built around real communities, nfts, friendships, chaos, conviction, and collective belief during the hardest times.
Not just perps, casinos, AI agents, and whatever the next shiny meta is.
Maybe this is simply what happens when ecosystems mature. Maybe I'm just too washed now.
I genuinely miss when the people themselves felt like the soul of Solana.
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
Jupiter Spend is legit evolving into a true fintech bank.
You can now do this straight from Jupiter Mobile:
→ Spend anywhere Visa is accepted worldwide
→ Non-custodial, on-chain USDC settlement
→ Send money internationally in fiat
→ Earn cashback 4-10% on purchases
→ Apple Pay & Google Pay support
→ Low/no FX fees on USD transactions
→ Track your spending over time: daily, weekly, monthly.
→ See where you spend globally
→ Top merchants
→ Recurring payments
Pretty impressive considering it only launched a few months ago.
English
Use my referral link to join the JTX waitlist. jtx.com/?ref=giancarlo…
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
Solana has taken the hard path, more than any other crypto network.
@anza_xyz has spent years doing the difficult engineering work required to push network performance closer and closer to centralized systems, while still preserving a credibly neutral, globally decentralized network.
And it’s not just Anza. Eco teams like @Helius, @Rakurai_io, @triton_one, @jito_sol and many others have continuously made improvements across the stack: validators, clients, networking, block propagation, RPC infra, MEV systems, tooling, and dev experience.
That work compounds. And over time, it has made @Solana increasingly competitive with centralized alternatives without abandoning the core properties that make crypto matter in the first place.
Meanwhile, many newer systems like L2s, appchains, Hyperliquid-style architectures, or new corpo chains have taken shortcuts (or have just stagnated completely like Ethereum L1 or Bitcoin).
The real question is whether maintaining decentralization actually matters from both an economic and regulatory standpoint.
I strongly believe it does long-term.
I believe in @toly’s vision of a global financial network that cannot be shut down, is open to anyone, and allows assets of all kinds to find price discovery in a far more efficient and permissionless way than today’s financial system.
That vision is much harder to build.
But if crypto succeeds at scale, I believe the world ultimately converges in that direction.
Hats off to everyone working at Solana’s infra layer.
Brennan Watt@bw_solana
200ms slots are on the menu… … Agave v4.2 is going to be the most insane client upgrade in Solana history
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
This is an extraordinary document written by the research arm of China's spy agency (the powerful MSS, basically the CIA and the FBI all wrapped in one) that absolutely zero media has picked up on.
As far as I can see, I'm the first person to write about it even though it was published (in Chinese) on May 13th on chinadiplomacy.org.cn, a website of China's Ministry of Foreign Affairs.
The document contains perhaps the most authoritative description of where China thinks its relationship with the U.S. stands, and where it’s headed.
The title of the report is “The Great Global Transformation and the Path to U.S.–China Coexistence” and I provide a full translation of it in my article, the link of which is at the bottom of this post.
To summarize briefly the most important - and, perhaps, surprising - aspect of the document: China's spy agency - the one institution whose entire job is to worry about the U.S. threat - has largely stopped worrying.
That's really what transpires from the document. They use a strategic framework borrowed from Mao's "protracted war" theory and, according to this framework, America's offensive phase is finished and China weathered the storm intact.
The question is no longer "how do we survive America?" but "how do we manage America?" - and they're proposing a six-step relationship recovery program.
I'll let you read the full document as well as my analysis of it here: open.substack.com/pub/arnaudbert…
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
Been waiting on this for a while. Going to abuse it for DCA.
> Set your bins to the price range you want
> Fills don’t sell back
> Earn fees while you DCA
> Basically 0 slippage
Picking your shape when buying:
> Bid-ask: loads up more at the bottom
> Spot: even distribution across the range
> Curve: loads up more at the top
Meteora@MeteoraAG
Meteora has been the edge for LPs on @solana. Today, we become the edge for traders. Introducing Limit Orders: get paid when you trade.
English
Learned this after the 3rd 10 year old software refresh as a contractor. The external dependencies are the ones that hurt the most to upgrade. Internalizing things especially for the smaller teams and with AI tools.
Mitchell Hashimoto@mitchellh
Fork your dependencies, trim them to only your use case, never update unless it breaks for your users. I’ve been vocal about this for 10+ years. I’ve always said that updating is way riskier than latent bugs (which can be tracked and CVEs monitored). If you are updating a dependency, it’s on you to analyze every single commit in the full transitive set of dependencies. If you dont see anything compelling, dont update! I remember at HashiCorp once in awhile an engineer would try to update a dep or replace a DIY lib with an external one and id always ask “show me the commit we need.” Dont update for the sake of it. Feeling pretty swell about this mentality with all the supply chain attacks happening.
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
Fork your dependencies, trim them to only your use case, never update unless it breaks for your users. I’ve been vocal about this for 10+ years. I’ve always said that updating is way riskier than latent bugs (which can be tracked and CVEs monitored).
If you are updating a dependency, it’s on you to analyze every single commit in the full transitive set of dependencies. If you dont see anything compelling, dont update!
I remember at HashiCorp once in awhile an engineer would try to update a dep or replace a DIY lib with an external one and id always ask “show me the commit we need.” Dont update for the sake of it.
Feeling pretty swell about this mentality with all the supply chain attacks happening.
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
The State of Solana
TL;DR
$SOL down 33% QoQ to $83.
On-chain activity hit new highs: 112.6M daily non-vote transactions, up 50% QoQ.
Solana dApp revenue remained high at $342M despite the bear market.
PumpFun, Axiom, Phantom and Jupiter generated the most revenue.
RWAs grew 43%.
DeFi TVL down 22% mostly due to SOL price action.
Stablecoin market cap new ATH at $15B.
Payments and institutional adoption expanded, with activity involving firms such as Visa, Stripe, Worldpay, Western Union, PayPal, Mastercard, and others.
AI agents starting to appear, x402 adoption expanded.
Alpenglow is the major upcoming upgrade, aiming to reduce transaction finality from around 12.8 seconds to roughly 150 milliseconds.
Solana@solana
Messari State of Solana Q1 2026 report is now live TLDR? RWAs up +43% to $2B and Solana now settles nearly half of stablecoin volume across major networks 🔥
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
In this video, @pmarca describes how some people have stopped sleeping because they code with AI. His model is that the ‘opportunity cost’ of sleeping is too high, meaning that they can produce so much value with AI that they’d better not sleep.
My model of what is happening is a novelty effect mixed with gamification. There are people who are naturally drawn to new experiences and tend to embrace them too much. Also, building with AI is a bit of a game. There is a Minecraft feel to it: you quickly build a house and then walls around the house and so forth.
I’ve seen it with some graduate students who produce massive amounts of ‘stuff’. Some of them become convinced that they are the new Einstein.
They are not.
What we have is a genuine glut of ‘research outputs’.
We have been there before, repeatedly. For example, about 20 years ago, blogging became a ‘thing’. The WordPress empire came to be. Everyone and their dog had a blog. The lady running my university started a travel blog soon after she had to leave due to her general incompetence. I don’t expect that her career as a blogger lasted very long.
I suspect that the same will apply with AI. Throwing stuff at AI is fun. Building valuable software is difficult.
Even though we all have access to a word processor, few of us become novelists.
AI will not turn most of us into software architects.
What is happening is Cargo Cult Software development. Cargo cultism is a reference to Pacific Islanders who would copy the US military after the US was gone, in the hope of getting the goods that the US was providing.
Software is not done for. We just have a new tool. Building useful software is still hard. Some people like it, but most will tire of it and find faster tricks. People will add skills or other lightweight techniques.
There will not be a world where everyone writes their own operating system or browser.
Uncle Bob Martin@unclebobmartin
Yup.
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
☁️ All SOL should be liquid staked
And after outperforming main competitors by 20%+ in 2025, INF remains the standout LST, averaging ~6.3% since our v2 upgrade.
Get started 👇
app.sanctum.so/infinity
English
giancarloangulo IBRL 🇵🇭☁️🔥💃 retweetledi
The public internet itself was the ceiling on @Solana network optimization.
Until two solutions were launched in the last two months:
RPC 2.0, by @triton_one.
Edge, by @DoubleZero.
Read more: ↓
Triton One 🌊🌋@triton_one
Three of Solana's core foundations are being rebuilt in 2026: consensus (Alpenglow, by @anza_xyz), reads (RPC 2.0, by us), and network (Edge, by @doublezero). For years, the public internet was the ceiling on every network optimisation. DoubleZero first replaced it on the write side, providing validators with a dedicated fibre path for receiving transactions. Edge extends the same idea outward: validators publish shreds into the DoubleZero network, multicast distributes them by physical distance, and switching hardware replicates them at the last hop, allowing the network to scale without slowing down. Our new blog post with @doublezero goes deeper into how it works and what it opens up for the ecosystem: blog.triton.one/rethinking-the…
English