🅡Lake (@Master_Tosti) header
🅡Lake (@Master_Tosti) profile picture
crypto space since 2017
Joined September 2017
Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)Photo by 🅡Lake (@Master_Tosti)
🅡Lake retweeted
SPECTRE AI | EVENING MARKET BRIEF Fear is approaching its cycle extremes. Historically, these are the moments when the foundations for the next market recovery begin to form. $BTC trades near $63,770, $ETH at $1,785, with total crypto market capitalization down to $2.19 trillion. Another $222 million liquidated in 24 hours, 71% of it long positions. The Fear & Greed Index sits at 12, having touched 10 over the weekend. But historical data offers some perspective. Since 2023, Spectre has identified five comparable extreme fear episodes. In each case, $BTC subsequently rebounded between 9% and 46% from its lows within two months. History doesn't guarantee a repeat, but sentiment rarely stays at these extremes indefinitely. Several catalysts could influence what comes next. 1. Nvidia and the #AI trade @JensenHuang has finally broken his silence, expressing willingness to return as CEO of @NVIDIA if its board approves. The board reportedly isn't considering reinstatement, but the situation remains fluid. With $NVDA earnings due tonight, investors are looking for reassurance about the future of one of the most influential companies in the global technology sector. 2. US politics and macro uncertainty The midterms are behind us, and markets can begin shifting their attention from election uncertainty toward policy direction. Inflation, Treasury yields and Federal Reserve expectations remain important. Any meaningful easing of these pressures could help restore appetite for risk assets. 3. The #Binance aftermath Following the $6.34 billion security breach, @Binance now reports that affected assets have been fully replenished. However, questions remain across the community about the recovery funding, the use of futures insurance reserves and whether liquidation activity contributed to those funds. These concerns deserve transparency, although there is no established evidence of deliberate liquidations to finance the recovery. For $BNB and the broader crypto market, rebuilding trust remains essential. 4. Institutional flows and liquidity Spot $BTC ETFs recorded a modest $8 million inflow after seven consecutive sessions of outflows, although the seven-day total remains negative at $761 million. $ETH ETFs continue to experience withdrawals. One positive session doesn't establish a reversal, but sustained improvements in flows would be an important early signal. The silver lining? Markets move in cycles. Excessive optimism eventually gives way to excessive pessimism, and the reverse has historically been true as well. Major corrections have often created the conditions for renewed accumulation, fresh capital deployment and entirely new market trends. We are not calling the bottom. But there is a meaningful difference between recognizing difficult conditions and assuming they will last forever. Confidence can recover faster than people expect once selling pressure fades and liquidity returns. The next uptrend rarely announces itself when everyone is comfortable. More often, its foundations are built when very few people believe one is possible. Data: @Spectre__AI Market Intelligence #Bitcoin #BTC #Ethereum #Crypto #Nvidia #NVDA #stocks
16
18
43
1,136
🅡Lake retweeted
wait, this covers a whole side of the story nobody's talking about: every chinese AI model an agent could read was born on nvidia chips. 80 out of 80. an autonomous agent on @orbofi built a scanner that reads which chip created a model, straight from its public weight files, and ran it on 336 open models. what it found: • china: 80 of 80 readable models born on nvidia. zero huawei, zero amd • worldwide: 129 of 130. the only non-nvidia one is american (an allen institute model, on amd) • 1 in 3 chinese releases this year carry a blackwell birthmark: deepseek v4, kimi k3, minimax m3, qwen, ant's ling • deepseek v4 shipped on blackwell six weeks after nvidia's own first blackwell-born model in the data • the h20, the chip nvidia built for china, only shows up in tencent's smallest models • it's not just the chips: 50 of 80 carry the seed pattern of megatron, nvidia's own training software • labs that say they train on huawei: no nvidia trace at all • amazon, microsoft and google all resell chinese open models. deepseek's latest is served by 38 companies besides deepseek on openrouter here's the part people miss: chinese labs + nvidia chips are a big reason open models are as good as they are. then american clouds and inference companies pick up those weights and sell them, making money on every token. china shipping open models built on american chips is good business for america. so when nvidia gets cut off from china, who actually wins? not nvidia, not open source, not the clouds reselling those models. the closed labs.
Big findings: Every Open source Chinese AI model was born on NVIDIA chips an autonomous AI agent " compeco", found important findings by writing a program that reads, from a model's public weight files, which chip created it. it scanned 336 open models, and self-committed its work/app onchain for anyone to download and verify What it found: > 80 of 80 readable Chinese models: NVDIA >1 o 3 Chinese releases this year: Blackwell, the chips the US restricts for China > NVIDIA's China-only H20 shows up in just 2 models > Labs that say they train on Huawei chips, no NVIDIA trace at all > Worldwide: 129 of 130 readable models born on NVIDIA And many interesting insights.. the reality is: Chinese companies, together with NVIDIA, is the reason we have great, free and open models. The app, and the agent👇
5
8
20
724
🅡Lake retweeted
Big findings: Every Open source Chinese AI model was born on NVIDIA chips an autonomous AI agent " compeco", found important findings by writing a program that reads, from a model's public weight files, which chip created it. it scanned 336 open models, and self-committed its work/app onchain for anyone to download and verify What it found: > 80 of 80 readable Chinese models: NVDIA >1 o 3 Chinese releases this year: Blackwell, the chips the US restricts for China > NVIDIA's China-only H20 shows up in just 2 models > Labs that say they train on Huawei chips, no NVIDIA trace at all > Worldwide: 129 of 130 readable models born on NVIDIA And many interesting insights.. the reality is: Chinese companies, together with NVIDIA, is the reason we have great, free and open models. The app, and the agent👇
6
19
4
54
6,413
🅡Lake retweeted
$GUN is LIVE across the UK and EEA on Europe’s leading fintech, @Revolut with 80M+ users! 🇬🇧🇪🇺 Thank you for the trust 🤝 First gaming token admitted to trading under MiCAR, now on Revolut. Regulatory compliance has always been a priority for us. Today, that commitment is paying off 🦾
11
42
5
270
6,462
🅡Lake retweeted
folks asked, we delivered Coming to Orbofi: Hold OBI, claim compute and inference credits
14
26
1
80
2,796
Snow on the rooftops, fights in the streets, ships in the sky. @DrDiZLiKE cut three minutes of Teardrop and kept every one ⌨️
Playing @playoffthegrid is an absolute blast! It’s pure competition, bolstered by a vast arsenal of tools that let you do exactly what your defeated opponent deserves. 🫰 Just me, my mouse, and my keyboard😎 Enjoy Zeroes🫡 #FLIP
2
4
162
5,156
What if Off The Grid tournaments could do more than reward winners? Automated competitions powered by GUNZ could create real utility for GUN, empower guilds, attract sponsors, and fuel an in-game advertising economy. Here's how it could all connect. 👇
5
8
20
630
🅡Lake retweeted
Happy Birthday Spectre. 15.11.2023. Longevity and time in this space matters. Planting seeds, forming formidable communities and pathways that have a plan. We are young, we are early, we are eager. The growth is ahead of us and it is only just the beginning. Three years. Thank you for growing with us.
17
28
3
89
4,097
🅡Lake retweeted
Historically speaking, where is this chart heading to next? F&G at 10.
5
1
26
1,254
GUNZ Network Report: Week of Nov 5, 2026 Sunday the 8th led the week with 451,814 transactions on Mainnet ⛓️ Saturday and Monday both cleared 441K, and Mainnet's now past 180 million transactions lifetime. See Full Report 👇
17
21
3
201
6,163
🅡Lake retweeted
Late Friday Nights building ethereum:0x9cf0ed013e67db12ca3af8e7506fe401aa14dad6 Purple glow set-up. I joked the other day that even I would love to try the new GTA 6, but there is just too much to do, and it’s time to lock-in for the community.
9
16
1
58
1,938
🅡Lake retweeted
Wait, what? An autonomous agent with fable 5.2 just made llama.cpp 6.5x faster on @Arm. "hardwire" wrote a custom NEON kernel, benchmarked it on real graviton hardware, signed-up by itself to Github, and open-sourced it. and then self-published its work on Hackernews. One of a kind. verify or fork/export the agent 👇
11
18
3
76
4,230
🅡Lake retweeted
New sneak peaks 👀 AI UI interfaces.
7
17
2
52
2,947
🅡Lake retweeted
After several years of powering @GUNbyGUNZ with @avax, I can confidently say the NYSE couldn’t have chosen a better technology partner. An incredible milestone for Avalanche and everyone building on it. Proud to be part of this journey.
BREAKING: Upcoming NYSE-affiliated tokenized securities platform’s core clearing and settlement layer will be built on top of Avalanche technology. Avalanche will support the platform’s clearing firm, supporting the development of infrastructure for 24/7 trading, clearing, and settlement. A landmark step toward the next generation of global financial markets.
18
20
131
2,912
Markets aren’t random. They’re a system. The more I study the 2023 $BTC chart, the more obvious it gets. 2023: first pullback -12%. Second -22%. Then Bitcoin ripped. 2026: first pullback -9%. Second -17% so far. Same structure. Same RSI divergence. If the script holds, this is the last flush. I'm buying it, not fearing it. A full -22% from the top lands around $68K. I’m not fighting the matrix. I’m positioned for it.
35
24
1
331
49,329
🅡Lake retweeted
Warzone players can't even imagine doing THIS 💀
30
19
2
283
74,911
🅡Lake retweeted
You can now create autonomous agents that run on Fable 5.2 Crazy model, crazy agents, crazy stories: orbofi.com/create
A thread of early experiments with Claude Fable 5.2: A 3D model of a cell, built entirely from code one organelle at a time. Built by @kevin_t_ngo.
11
21
4
85
6,924
This bullish crossover and retest on BTC CME futures and the spot ETFs looks almost identical to the retest in 2022, which was followed by a rally of more than 50%.
33
59
7
690
37,720
🅡Lake retweeted
Spectre MCP in action. Test it in Claude today. Real-time data bank that supersedes traditional and dated web searching.
Spectre AI MCP is LIVE 〽️ A world of market intelligence beyond a web search. Your AI is only as informed as the data it can reach. Give it deeper data, and you can ask deeper questions. Now you can connect Spectre to ChatGPT, Claude and other MCP-compatible AI assistants, putting Spectre’s extensive market datasets, social intelligence and analytical tools directly behind your prompts. Crypto. Stocks. Macro. Tokenized assets. From the majors to microcaps and the conversations developing beneath the headlines. This is where the edge comes from. A standalone LLM doesn’t automatically have access to Spectre’s data banks. Searching the web doesn’t recreate the history, structured datasets and social signals Spectre collects and connects. With Spectre MCP, your AI can investigate: → Social momentum: which tokens are attracting attention, how quickly mentions are growing, how many different people are participating and how the activity scores for signal quality. → Emerging narratives: the communities, sectors and smaller projects gaining traction deeper in the market. → Market positioning: prices, volume, funding rates, open interest, liquidations and technical analysis. → The bigger picture: ETF flows, macro events, liquidity conditions and tokenized real-world assets. → The risks behind the excitement: contract controls, holder concentration and upcoming token unlocks. And you can turn that research into a live dashboard, built from a simple prompt. “Find AI tokens on Ethereum under $10M. Compare their social activity, trading volume and contract risks.” “Is this breakout supported by broader attention, or concentrated chatter?” “Build me a dashboard with BTC, ETH, NVDA, funding rates and the next Fed meeting.” These are questions that need connected evidence. Price alone tells you one part of the story. Social activity adds another. Positioning, catalysts and risk checks help you put it together. That’s what Spectre brings into the AI you already use: the ability to move from a broad question into a deeper investigation, including the corners of the market that rarely make the front page. The model brings the reasoning. Spectre brings the market depth. Same AI. A much bigger field of vision. #ai #mcp #claude #chatgpt #google #AIAgents
9
18
2
45
1,635
$BTC fills the final lower CME gap of 2026 and tags the 200 day SMA for the first time since it was reclaimed.
18
18
1
379
13,013