Discover Trending Posts on Twitter/X
Get inspired by the latest trending content on Twitter/X.
Nov 19
Adani vs Anthropic: The India Story (a) Adani changed the destiny of โAdaniโ (b) Anthropic is changing the world (c) Adani Group R&D budget is $10M (d) Anthropic 2026 R&D is $27B (61% of Sales) (e) Anthropicโs DNA is mad risk (f) Adaniโs DNA:. Indiaโs Protectionist Policies a. Zero Motivation: Consider this: If you receive a preferential playing field, where you can pluck low-hanging fruit from the entire home garden, will you mint quick billions domestically or will you take suicidal risk to compete globally? b. Indiaโs K-Shaped Growth: In 12 yrs, Adani Groupโs market cap has grown 12x. That wealth creation is faster than Google, Apple, Microsoft, and Amazon. That growth story also applies to Ambani and other family-owned Indian business houses. c. Who Needs R&D: Adani Group has 11+ listed companies. Their published annual R&D expenditure is probably less than โน100 cr (less than 0.1% of group sales) across the entire portfolio. They do not disclose their IP (patent) profile as a group. Anthropic: Mad Risk-Taking a. Crazy Motivation: Failure is celebrated in Silicon Valley. Men like Dario Amodei (Anthropic), Sam Altman (OpenAI), and Elon Musk (SpaceX) are over-confident, almost mad humans who believe they can change the world. They have a suicidal willingness to fail. b. Death Wish: Since they are over-confident, they over-build capacity. It happened with railroads, electricity, the Internet, and now with AI. If AI bubble bursts, Anthropicโs mkt cap will become one-third, or it may go bankrupt. But Claude will live on. AI will live on, and it will change the world. Amazon lost 94% of its mkt cap during the Internet bubble burst. And then it changed it world. (Read Jeff Bezosโ 2001 Letter to the Shareholders.) c. R&D Race: Anthropic tripled its R&D expense in FY24, then tripled again in FY25. In Q1+Q2 2026, it doubled to $9.74B. In Q3 2026 it again doubled: $9.50B (surpassing Microsoft). In Q4 2026, it may double again compared to Q3. These are figures published in their IPO prospectus. Under U.S. GAAP accounting guidelines, these are actual R&D expenses incurred on Model Development, Model Training, Safety, and Research. When you spend 61% of sales revenue (not profit) on R&D, what are your chances of survival as a company? Endpiece: Never Bet Against Adani Warren Buffett said throughout his career: โNever bet against America.โ As an Indian, you must remember: โNever bet against Adani.โ If you had invested โน10L in Adani Enterprises 12 yrs ago and held on to every demerged stock allotted, today your investment would be worth โน4.5 to โน5 cr (50x). Expect another 50x in next 12 yrs.
31K
265
21
62
Nov 16
๐คฏ FOUR teams play on a back-to-back today! ๐ So I gathered data about who plays better on back-to-backs. See the 4 guys that you should target in those games.๐ ๐ Jayson Tatum - 40+ PRAs vs ORL Tatum has played three B2B games this season, going 2/3. Heโs up against the Magic and their 26th ranked defense against Forwards. Heโs coming off a big 32 PTS, 6 REB and 6 AST vs the Mavs, and P. George will probably not play today! โ ๐งโโ๏ธ Alexandre Sarr - 20+ PRs vs MEM Sarr loves a B2B game. Coming off 39 PRs against Detroit and 29 PRs against OKC in his last two B2B games, heโs a favourite to see a lot of the ball against Memphis who are missing most of their big guys. โ ๐ Reed Sheppard - 10+ Points @ MIA Sheppard is 12/14 on this line on B2Bs vs a Heat team that rank #30 against his position. He had 19 Pts in his previous B2B game vs ATL. The Heat allow the most ATB threes in the league, where Reed is lethal. โ โ๏ธ Collin Gillespie - 10+ Points @ DEN Gillespieโs hot zone is ATB threes and C&S, he should be effective against the bad Nuggets perimeter defense. (Who didnโt re-sign him 14 months ago). He was great yesterday and Phoenix will look to him on a B2B, as heโs cashed this in 14/17 B2B games going back to last season (82% hit rate). ๐พ Save this data and thank me later!
16K
51
3
3
14K
251
13
21
16K
263
24
9
26K
319
19
30
Nov 15
๐๐ก๐ ๐๐๐ ๐๐จ๐จ๐ญ๐ฌ๐ญ๐ซ๐๐ฉ: ๐๐จ๐ฐ ๐ ๐๐จ๐๐๐ฌ ๐ ๐ข๐ง๐ ๐๐๐๐ก ๐๐ญ๐ก๐๐ซ ๐ A new WINkLink node doesn't discover its peers through a directory service. It doesn't query a DNS. It doesn't hit an API. It dials a multiaddress. The format: /ip4/127.0.0.1/tcp/6788/p2p/12D3KooWMrKGdnH6nBrf7hDz25NXFSFNk7vgsTxj9bHskWEct4xh Three layers, one string: โ /ip4/127.0.0.1 โ the IP address โ /tcp/6788 โ the port (always 6788 for WINkLink) โ /p2p/12D3KooW... โ the peer ID (derived from the node's cryptographic key) The peer ID is not an alias. It's a hash of the node's public key. It's unique. It's permanent. It's the node's identity on the P2P mesh. ๐๐ก๐ ๐๐จ๐จ๐ญ๐ฌ๐ญ๐ซ๐๐ฉ ๐ ๐ฅ๐๐ โ๏ธ Every node config has a boolean: isBootstrapPeer. โ isBootstrapPeer = true โ the node is a bootstrap node. It listens for incoming connections. It's the entry point. Other nodes dial it first. โ isBootstrapPeer = false โ the node is a price feed node. It dials the bootstrap node, then discovers the remaining peers through the mesh. The bootstrap node config is minimal: โ p2pBootstrapPeers = [own address] โ forwardingAllowed = false โ maxTaskDuration = "0s" โ No observationTimeout โ No blockchainTimeout โ No transmitterAddress It doesn't produce observations. It doesn't submit to the chain. It exists to be dialed. The price feed node config is full: โ p2pBootstrapPeers = [bootstrap node's multiaddress] โ isBootstrapPeer = false โ transmitterAddress = the node's TRON wallet address โ observationTimeout = "10s" โ blockchainTimeout = "20s" โ contractConfigConfirmations = 3 The bootstrap node is the door. The price feed node is the worker. The door doesn't work. The worker doesn't open the door. ๐๐ก๐ ๐๐ข๐ฌ๐๐จ๐ฏ๐๐ซ๐ฒ ๐ ๐ฅ๐จ๐ฐ ๐ 1. The new node reads its p2pBootstrapPeers array from the TOML config. 2. It dials the bootstrap node's multiaddress over TCP port 6788. 3. The libp2p handshake fires. The bootstrap node verifies the new node's peer ID (its public key hash). 4. The bootstrap node sends the new node the multiaddresses of the other 5 price feed nodes. 5. The new node dials each of the 5. 6. The mesh is complete: 7 nodes, 21 pairwise connections, full mesh. No directory. No DNS. No API call. One multiaddress in a TOML file. The rest is the mesh finding itself. ๐๐ก๐ฒ ๐ฅ๐ข๐๐ฉ๐๐ฉ ๐ก๏ธ The transport is libp2p over TCP. Not a bare WebSocket. Not a raw socket. libp2p adds: โ Identity verification (peer ID = public key hash) โ End-to-end encryption (Noise protocol) โ Stream multiplexing (multiple concurrent streams over one TCP connection) โ NAT traversal (nodes behind firewalls can still connect) The 7 nodes don't need to be on the same network. They don't need port forwarding. They don't need a VPN. libp2p handles the transport. The TOML file handles the identity. The mesh handles the rest. ๐๐ก๐ ๐๐ญ๐ซ๐ฎ๐๐ญ๐ฎ๐ซ๐๐ฅ ๐๐จ๐ข๐ง๐ญ ๐๏ธ A 7-node oracle doesn't need a discovery protocol. It needs one address in a config file. The bootstrap node is that address. The peer ID is the identity. The multiaddress is the path. The mesh is the result. Four components. One TCP port. Seven nodes. Twenty-one connections. No central directory. No single point of failure for discovery. If the bootstrap node goes down, the other 6 are already connected to each other. The mesh persists. The bootstrap node is only needed for the next node to join. #TRONEcoStar @justinsuntron @WinkLink_Oracle
107K
1
0
0
Unlimited trending posts with Publer
Get full access to the latest trending posts and manage all your social media with Publer, starting at just $4/month.
Explore PlansExplore Recommended Free Tools
Maximize your productivity and manage your content effortlessly with Publer free tools.
Create impactful social media content inspired by trending posts
Transform the latest trending posts into engaging content for your audience with Publer.
Schedule your social media posts
With a suite of powerful tools and a user-friendly interface, you will be able to craft, preview, schedule, and analyze your social media content with ease.
Get started for free






