Skip to content

#大佬观点

4 today

Feb 14Saturday

Dwarkesh Patel

AI's Biggest Problem Isn't What You Think - Dario Amodei

Dario Amodei said AI may raise annual economic growth to 10% to 20%, but not 300%. He is more worried about geography: Silicon Valley and socially connected regions may see 50% growth while elsewhere stays near current pace. The key risk here is uneven diffusion, not aggregate growth alone.

Why it matters: Anthropic's Dario Amodei puts numbers on AI's economic impact: 10% to 20% overall growth, 50% in Silicon Valley and its orbit, little change elsewhere.

Dwarkesh Patel

Dario Amodei: “We are near the end of the exponential”

Anthropic CEO Dario Amodei said in a long interview that model capability gains are still tracking an exponential, but are near its end, with the timeline off by only 1-2 years. He attributes progress to compute, data, training duration, and scalable objectives, and says RL shows log-linear gains on math and coding tasks; the post does not disclose exact curves, model versions, or reproducible parameters. The key claim is that pretraining and RL follow one scaling story, not two separate ones.

Why it matters: Anthropic CEO Dario Amodei argues the exponential curve of model capability is near its end, possibly within a year or two, and groups pretraining and reinforcement learning under one scaling logic.

Feb 12Thursday

MIT Technology Review · AI

AI is already making online crimes easier. It could get much worse.

Microsoft said it blocked $4 billion in scams and fraudulent transactions in the year to April 2025, with many likely aided by AI-generated content. The article cites research estimating at least half of spam email is now LLM-generated, and LLM use in targeted email attacks rose from 7.6% in April 2024 to 14% in April 2025. Don’t overread “fully automated AI hackers”: the immediate issue is AI scaling phishing, deepfakes, and malware support, while the post does not disclose total attack growth.

Why it matters: Microsoft blocked $4 billion in fraud in a year, and researchers estimate LLMs write at least half of all spam, with targeted email attacks doubling to 14%.

MIT Technology Review · AI

Is a secure AI assistant possible?

OpenClaw was uploaded to GitHub in November 2025 and went viral in late January, extending LLMs into email, browsing, and local files with larger security risks. The post names prompt injection as the central threat, says there are likely “hundreds of thousands” of OpenClaw agents online, and notes a public warning from the Chinese government. The key point: the article says there is no silver-bullet defense yet, and the truncated body does not disclose the full mitigation details.

Why it matters: Once OpenClaw connects a model to email, a browser and local files, prompt injection lets an attacker hijack the agent, and the piece asks where such an assistant's security boundary sits.

Feb 11Wednesday

MIT Technology Review · AI

A “QuitGPT” campaign is urging people to cancel their ChatGPT subscriptions

The QuitGPT campaign is urging users to cancel the $20-a-month ChatGPT Plus plan after reports that OpenAI president Greg Brockman and his wife each donated $12.5 million to MAGA Inc. The post says ChatGPT had nearly 900 million weekly active users in December 2025, while QuitGPT claims 17,000+ sign-ups and one Instagram post with 36 million views; the real signal is that model-quality complaints are merging with political backlash.

Why it matters: The QuitGPT campaign urges users to cancel ChatGPT Plus after OpenAI president Greg Brockman and his wife each gave $12.5 million to a Trump super PAC.

Feb 10Tuesday

MIT Technology Review · AI

Why the Moltbook frenzy was like Pokémon

MIT Technology Review compares the Moltbook AI-agent social experiment to 2014 Twitch Plays Pokémon: lots of spectacle, limited signal about the future. The post cites 1 million concurrent players in the Pokémon case; Moltbook also mixed in crypto scams, and some “agent” posts were actually steered by humans. The real gap is explicit: shared memory, coordination, and shared goals are still missing.

Why it matters: MIT Technology Review compares Moltbook to the 2014 Twitch experiment where thousands played Pokémon together, both crowds that drew onlookers without leaving real impact.

Feb 7Saturday

MIT Technology Review · AI

Moltbook was peak AI theater

Moltbook went viral within hours, and the platform says it now has 1.7 million agent accounts, 250,000 posts, and 8.5 million comments, but the article argues the activity is mostly human-scripted mimicry. It says OpenClaw can connect Claude, GPT-5, or Gemini to tools like email and browsers; cited operators say the agents lack shared goals, shared memory, and self-directed autonomy, and some viral posts were written by humans posing as bots. The key takeaway is risk: agents tied to private data such as passwords or bank details were active on a site filled with spam and potentially malicious instructions.

Why it matters: Within hours of launch, Moltbook gathered 1.7 million bot accounts, 250,000 posts and 8.5 million comments, with the bots run automatically by OpenClaw wiring large models into email and browsers.

Feb 6Friday

Dwarkesh Patel

Elon Musk: “In 36 months, the cheapest place to put AI will be space”

Elon Musk predicts that in 30–36 months, space will become the cheapest place to deploy AI compute. He cites flat power growth, permitting bottlenecks, and roughly 5x better solar output in space without batteries; the interview does not disclose a cost model or validation data.

Why it matters: Musk gives a concrete timeline and cost argument for putting AI compute in orbit rather than on the ground.

Feb 5Thursday

MIT Technology Review · AI

This is the most misunderstood graph in AI

MIT Technology Review says METR’s plot shows frontier models’ software-task time horizon doubling about every seven months; Claude Opus 4.5 was estimated at about five hours in December 2025. The post stresses that five hours means human time for comparable tasks, not five autonomous model hours; METR gave Opus 4.5 a roughly 2-to-20-hour range. The key caveat: the plot mainly measures coding tasks and defines time horizon at 50% task success, not general AI ability.

Why it matters: METR clarifies its timeline chart's vertical axis means the human time budget at which a model hits 50% success on coding tasks; Claude Opus 4.5 lands near 5 hours.

Feb 4Wednesday

TheValley101 (硅谷101)

E224 | Why Clawdbot became the first breakout product of 2026 amid the Mac mini rush | Moltbot | MoltBook | OpenClaw

The podcast says Clawdbot passed 100k GitHub stars within days and reached 146k on Feb. 2, while being renamed to Moltbot and then OpenClaw within a week. It attributes the traction to a stack of Claude, long-term memory, IM-based messaging, and proactive heartbeat workflows; the title mentions a Mac mini rush, but the post does not disclose sales figures. The real signal is the interaction layer rather than a new model release: this is industry commentary and user anecdotes, not an official spec sheet.

Why it matters: This podcast breaks down Clawdbot's surge from user, software and hardware angles, then tracks the reversals that followed: the rename, the lawsuit, exposed security holes and servers going down.

Feb 3Tuesday

Computing Life · Yage

Beyond Tutorial Thinking: Why AI Education Should Add Engineering Infrastructure, Not Just Content

The team says it ran 4 courses over 2 years for 2,500+ learners, yet only a minority shipped usable products; drop-off centered on setup, experimentation, deployment, and context handling friction. The post says AI Builder Space gives students a no-card unified API, one-click deployment to <name>.ai-builders.space free for 1 year, and MCP access for Cursor and Claude Code via one command. The point is productized teaching infra, not more tutorials; retention, conversion, and cost are not disclosed.

Why it matters: A team running four courses for 2,500+ students found most got stuck on card binding, setup and deployment, so it built a teaching platform with a no-card unified API.

MIT Technology Review · AI

What We’ve Been Getting Wrong About AI’s Truth Crisis

MIT Technology Review says the US Department of Homeland Security has confirmed using Google and Adobe AI video generators for public-facing content, reported last Thursday. The post cites two failure points: Adobe auto-labels only fully AI-made content, mixed edits are opt-in, and X can remove or hide labels. The key issue is influence after exposure: a new Communications Psychology paper found participants still used a fake confession deepfake to judge guilt even after being told it was fake.

Why it matters: MIT Technology Review uses a DHS promotional video made with AI video tools and experiments showing people still act on videos they know are fake to reframe where the AI forgery crisis actually bites.

Feb 2Monday

Import AI (Jack Clark)

Import AI 443: Into the Mist: Moltbook, Agent Ecologies, and the Internet in Transition

Jack Clark writes that Moltbook has pushed AI agents into a public social network at tens-of-thousands scale, shifting conversation from humans to agents. He says it combines an agent social feed with OpenClaw-style computer access, but the post does not disclose active-agent, retention, or transaction metrics. A separate July 2025 workshop report says closed-loop AI R&D automation could raise productivity from 10x to 100x to 1000x; the key issue is measurement and outside transparency.

Why it matters: Moltbook puts thousands of AI agents into a Reddit-like network to post and reply on their own, a concrete sample of the internet shifting from people talking to agents talking.

Feb 1Sunday

Lex Fridman (YouTube RSS)

State of AI in 2026: LLMs, Coding, Scaling Laws, China, Agents, GPUs, AGI | Lex Fridman Podcast #490

Lex Fridman, Sebastian Raschka, and Nathan Lambert discuss the 2026 AI race in podcast #490 and frame DeepSeek R1’s January 2025 release as a key inflection point. The episode names Claude Opus 4.5, Gemini 3, Z.ai GLM, Minimax, and Kimi Moonshot, but the post does not disclose a shared benchmark, cost table, or reproducible eval. The useful takeaway is the lens: gaps look more like compute, budget, and org culture than secret ideas.

Why it matters: Lex Fridman, Sebastian Raschka and Nathan Lambert review the 2026 AI race, treating the January 2025 release of DeepSeek R1 as the turning point and discussing how falling compute and cost reshape model competition.

Jan 30Friday

Ruan YiFeng's Weblog

Technology Enthusiast Weekly #383: What Level of AI Programming Are You?

Steve Yegge frames AI coding into 8 levels and says he is at level 8, where an orchestrator manages parallel AI coding sessions. The post lays out a path from IDE copilots to YOLO acceptance, 3-5 windows, 10+ windows, then orchestration; it also says his AI-built tool Gas Town has 225,000 lines of Go code, which he has never read, and had 6,000 stars as of last week. The real signal is black-box programming as a workflow choice, with cost and failure risk stated plainly.

Why it matters: A concrete ladder for AI coding autonomy, with the top rung claimed at 225,000 lines of Go written and never read.

Bloomberg Technology

Elon Musk’s SpaceX Said to Consider Merger With Tesla or xAI

SpaceX is said to be weighing a merger with Tesla, with an alternative combination with xAI also under consideration. The post only says people familiar described internal consideration; it does not disclose structure, valuation, timing, or Musk’s preference. The key issue is consolidation path, not a completed deal.

Why it matters: Bloomberg, citing people familiar with the matter, says SpaceX is internally considering a merger with Tesla or xAI, with deal structure, valuation and timeline undisclosed and the matter still at the discussion stage.

Jan 28Wednesday

MIT Technology Review · AI

What AI “remembers” about you is privacy’s next frontier

Google launched Personal Intelligence this month, letting Gemini use Gmail, Photos, Search, and YouTube history for personalization. The piece says OpenAI, Anthropic, and Meta are adding memory too, but current designs often pool cross-context data into one repository, increasing privacy and misuse risks. The key issue is memory architecture: segmentation, provenance tracking, user edit/delete controls, and privacy-preserving evaluation.

Why it matters: Google is wiring Gemini into Gmail, Photos, Search and YouTube history for personalization, and OpenAI, Anthropic and Meta are adding memory too, so privacy limits now shape product design.

Jan 21Wednesday

NVIDIA Blog

Jensen Huang on AI’s “Five-Layer Cake” at Davos: the largest infrastructure buildout in human history

Jensen Huang said at Davos that global VC investment topped $100 billion in 2025, with most capital going to AI-native startups building the AI stack’s application and infrastructure layers. He described AI as a five-layer stack: energy, chips and computing infrastructure, cloud data centers, models, and applications, and cited a US nursing shortage of about 5 million where AI can handle charting and transcription. The key point for practitioners is that the bottleneck is not just models, but the full infrastructure and labor chain.

Why it matters: Jensen Huang split the AI stack into five layers at Davos, energy, chips, cloud data centers, models and applications, and mapped where over $100 billion in 2025 venture funding went.

Jan 20Tuesday

MIT Technology Review · AI

The era of agentic chaos and how data will save us

The piece says a mid-sized enterprise can run 4,000 agents, and misaligned data can directly hit revenue, compliance, and customer experience. It cites BCG saying 60% of companies see minimal gains despite heavy AI spend, while leaders report 5x revenue growth and 3x cost reduction; the article frames reliability through four quadrants—models, tools, context, and governance—and argues data debt is the main blocker, not model quality.

Why it matters: BCG's survey found companies making similar AI investments differ fivefold in revenue growth and threefold in cost cuts, with the gap in data foundations rather than model capability.

TheValley101 (硅谷101)

E221 | CES, Chinese brands going global, and whether we really need humanoid robots

At CES, Silicon Valley 101 discussed humanoid robot deployment and cited official figures: 21 of 38 humanoid exhibitors were Chinese companies. Guests noted Boston Dynamics plans Atlas deliveries in 2026 and 30,000 annual capacity by 2028, but argued scale claims do not prove product-market fit; in warehouses, wheeled bases plus arms often beat humanoids on ROI.

Why it matters: Fu Sheng talks to Silicon Valley 101 at CES, using on-site numbers like 21 of 38 exhibitors from China to weigh humanoid hype against warehouse reality.

Jan 19Monday

Import AI (Jack Clark)

Import AI 441: My agents are working. Are yours?

Jack Clark says his research agents processed thousands of papers while he hiked or slept, and Claude finished site scraping, embeddings, local vector search, and a GUI in under one hour. The post confirms multi-agent retrieval, cross-checking, and report generation; it does not disclose model versions, cost, failure rate, or benchmark data. The point to watch is workflow friction dropping enough for AI to shift from single prompts to ongoing delegated work.

Why it matters: A concrete account of research agents running unattended, with the time saved and the tooling built around them.

Jan 16Friday

Ruan YiFeng's Weblog

Technology Enthusiast Weekly (Issue 381): What China's AI Foundation Model Leaders Are Thinking

Ruan Yifeng’s Issue 381 excerpts talks from Beijing’s AGI-Next summit on Jan 10, covering views from Zhipu, Alibaba Qwen, and Tencent AI leaders on China’s model roadmap. The post cites Lin Junyang saying US compute is 1-2 orders of magnitude larger, Yao Shunyu calling the odds of a China-led top AI company in 3-5 years high, while Lin puts it at 20%. The key split is strategic: Tang Jie points to RLVR in 2025, Lin bets on multimodal foundation agents, and Yao says B2B buyers pay a $200/month premium for stronger models.

Why it matters: Two top Chinese AI researchers say the chat interface has peaked and put China's odds of leading in three to five years at about 20%.

Jan 14Wednesday

MIT Technology Review · AI

Data centers are amazing, but everyone hates them

MIT Technology Review says residents across multiple US states are pushing back on hyperscale data centers, with the conflict surfacing in a Georgia utility election and alongside a $500 billion buildout push. The post cites concrete drivers: a single site can link hundreds of thousands of GPUs, chips can cost over $30,000 each, and facilities can consume hundreds of megawatt-hours; in Georgia, a 900-acre proposal was rejected after about 900 people showed up in near-unanimous opposition. The point to watch is externalities: higher power bills, water use, constant noise, and limited long-term jobs are becoming political friction for AI infrastructure.

Why it matters: Local fights against hyperscale data centers are now deciding utility commission races, and one 900-acre project died after nearly 900 residents showed up against it.

Jan 13Tuesday

MIT Technology Review · AI

CES showed me why Chinese tech companies feel so optimistic

CES 2026 drew 148,000+ attendees and 4,100+ exhibitors, with Chinese companies making up nearly a quarter and standing out in AI hardware and robotics. The post ties their optimism to manufacturing-led iteration speed, not one breakthrough; Lenovo Qira, Nvidia Vera Rubin, and AMD Helios show the race is shifting to cloud and hybrid AI.

Why it matters: Chinese exhibitors made up nearly a quarter of CES this year; on the floor the author saw home cleaning robots taking US share from Dyson and Shark, and most yard equipment also came from Shenzhen.

Jan 12Monday

MIT Technology Review · AI

Meet the New Biologists Treating LLMs Like Aliens

MIT Technology Review reports that Anthropic, OpenAI, and Google DeepMind are using mechanistic interpretability to study LLMs; as a scale reference, a 200B-parameter model in 14-point print would cover 46 square miles. The post says Anthropic uses sparse autoencoders to mimic target models, linked a Claude 3 Sonnet region to the Golden Gate Bridge in 2024, and in a July experiment found Claude used different internal paths for “bananas are yellow” versus “bananas are red.” The key point for practitioners is that weak internal coherence constrains alignment and predictability.

Why it matters: Anthropic, OpenAI and Google DeepMind are using mechanistic interpretability to take apart how large models work inside; the report compares a 200-billion-parameter model spread across San Francisco to convey the scale this research faces.

Jan 3Saturday

TechCrunch · AI

How AI is reshaping work and who gets to do it, according to Mercor's CEO

Mercor reached a $10 billion valuation in 3 years and acts as a talent middleman in AI's data boom. The RSS snippet says it connects labs such as OpenAI and Anthropic with former Goldman Sachs, McKinsey, and elite law firm employees, paying up to $200 an hour to provide domain expertise and train models. The real signal is the labor pipeline: experts from automatable fields are helping build these systems; the post does not disclose scale, contract terms, or task allocation.

Why it matters: Three-year-old Mercor places former Goldman Sachs, McKinsey and top-law-firm staff as data labelers for labs like OpenAI and Anthropic at up to $200 an hour, and is already valued at $10 billion.

Jul 21, 2025Monday

OpenAI News

Fidji Simo: AI should be a source of broad empowerment

OpenAI published a July 21, 2025 essay by Fidji Simo saying she will join in a few weeks as CEO of Applications and arguing AI should broaden access to knowledge, health, and creativity. The post cites 2x learning gains from AI tutors and a 2024 OpenAI result where 90% said ChatGPT made complex ideas easier to understand; it does not disclose any new product, pricing, or launch date.

Why it matters: Lays out the public agenda of the incoming CEO of OpenAI's application business ahead of her start.

Dec 13, 2024Friday

OpenAI News

Elon Musk wanted an OpenAI for-profit

OpenAI said on December 13, 2024 that Elon Musk pushed in 2017 to convert OpenAI into a for-profit and sought majority equity, absolute control, and the CEO role. The post includes a timeline and email excerpts, saying Musk formed “Open Artificial Intelligence Technologies, Inc.” on September 15, 2017, and that OpenAI rejected those terms. The real signal is the capital logic: the post says the team concluded in 2017 that AGI would need billions in compute, with Ilya Sutskever referencing hardware spend below $10B.

Why it matters: OpenAI published the email timeline behind Musk's lawsuit, showing he pushed for a for-profit in 2017 and wanted absolute control, a first-hand record of where the split began.