Industry News

Tracks company news, strategic changes, funding activities, and personnel adjustments across the blockchain and crypto industries, delivering a full-spectrum industry overview for our users.

After Half a Year as a Token Broker, She Has Fallen into Every Pitfall of the Relay Station Business

Sukie, who operated an AI API "middle station" service for six months, recently open-sourced her entire setup process. Her story reveals the harsh realities of this once-lucrative but now hyper-competitive market. The core challenge is cost. Legitimate, compliant API accounts are expensive. To compete, many players resort to cheaper, high-risk sources like stolen accounts. The market has seen prices plummet from 70-80% of official rates down to 30-50%, a level unsustainable for compliant operators. Sukie believes a 70-80% price point is the minimum for healthy margins using legitimate methods. A major mistake was targeting the Chinese market while incurring USD costs. She found Chinese developers extremely price-sensitive compared to Western clients, leading to thin margins compounded by currency and payment hurdles. Operational burdens are heavy: maintaining a pool of hundreds of accounts against rising platform bans, handling detailed technical support, and managing cross-border payments and invoices for different client types. Marketing channels like X (Twitter) and referrals work best, while platforms like Douyin (TikTok) and Xianyu have poor ROI due to low intent or pricing mismatches. The landscape shifted dramatically with high-profile entrants like Justin Sun, Fu Sheng, and the Trump family. For them, the middle station is a loss leader to attract users to their primary businesses—crypto ecosystems, corporate narratives, or token promotions. This makes competing on price alone impossible for independent operators. Sukie open-sourced her methodology both as marketing and to demystify the industry. By eliminating the "black box" technical premium, she hopes to shift competition from cutthroat pricing towards service quality, stability, and compliance. Her advice: this is not a viable full-time venture for newcomers. The compliant path can't compete with grey-area discounters or ecosystem-backed giants. If already involved, focus on niche B2B, academic, or overseas markets. The middle station business, she concludes, is an entry ticket, not a destination, in the broader AI landscape.

marsbit05/09 04:48

After Half a Year as a Token Broker, She Has Fallen into Every Pitfall of the Relay Station Business

marsbit05/09 04:48

a16z Weekly Chart: Tech Giants Rely on 'Side Hustle' Investments for Income, Great AI Products Can Sell Out in a Day

a16z Weekly Charts: Four Counterintuitive Signals in Tech 1. **Super Platforms' "Other Income"**: Amazon and Google recorded exceptionally high "other income" in Q1, largely from unrealized gains in their private investment portfolios (e.g., Amazon's Anthropic stake). This contributed to over one-third of their net profit, far above the historical 5-10%. The broader trend shows tech capital expenditure is now the primary driver of US GDP growth, accounting for 55% of all business investment. 2. **AI-Generated eBook Proliferation**: Since ChatGPT's launch, monthly eBook releases on Amazon have tripled to over 300,000, flooding the platform with AI-generated content. However, research indicates this has also increased the volume of "decent" books, providing a net gain in consumer surplus by 2025. AI tools have particularly boosted productivity for established authors. 3. **Call Center Jobs Defy AI Replacement**: Contrary to predictions, call center employment in the Philippines has grown steadily from 1.15 million in 2016 to 1.9 million in 2025, with further growth projected. In the US, customer service job postings are outperforming the overall market. The key reason: the full cost of voice AI agents remains roughly equal to human agents (~$92 vs. ~$90 per day). Cases like Klarna show initial replacement can lead to quality issues and re-hiring. 4. **Rapid Adoption of AI Mobile Apps**: AI app downloads, revenue, and user time spent on mobile nearly doubled year-over-year in Q1. The market is highly dynamic, with new products like Codex quickly surpassing incumbents like Claude Code in daily installs. In the B2B space, enterprises are using multiple AI vendors, with less than 20% relying on a single supplier, indicating no winner-takes-all dynamic yet.

marsbit05/09 04:38

a16z Weekly Chart: Tech Giants Rely on 'Side Hustle' Investments for Income, Great AI Products Can Sell Out in a Day

marsbit05/09 04:38

Musk vs. Altman: Who Will Be the 'Fisherman'?

Elon Musk and Sam Altman are locked in a fierce legal and commercial battle. Musk, a co-founder of OpenAI, has sued the company and Altman, alleging they betrayed its original non-profit, open-source mission by transforming into a for-profit entity with significant Microsoft backing, now valued at $852 billion. He demands damages, a return to a non-profit structure, and management changes. The lawsuit hinges on whether OpenAI's founding charter was a legally binding charitable trust or merely an idealistic statement. OpenAI counters that Musk himself pushed for a for-profit model in 2017 but left when he couldn't gain full control, and now acts as a commercial rival with his xAI venture. Despite the high-profile feud, the article suggests the real winners (the "fishermen") may be others in the AI race. While Musk has folded xAI into SpaceX to pursue a "space-based computing" vision, his Grok chatbot lags in market share and user growth compared to leaders. OpenAI faces its own challenges, notably from rival Anthropic, which is rapidly catching up in revenue and enterprise adoption. Musk is reportedly leasing significant computing power to Anthropic, creating an "enemy of my enemy" dynamic. Furthermore, Chinese AI models like DeepSeek are quickly closing the capability gap. Ultimately, the lawsuit is seen as setting a precedent for AI governance, but the intense competition between Musk and Altman may primarily benefit other players, infrastructure providers like Nvidia, and emerging third forces in the global AI landscape.

marsbit05/09 04:27

Musk vs. Altman: Who Will Be the 'Fisherman'?

marsbit05/09 04:27

Dissolving xAI, Musk Wants to Rebuild an AI Company Using Rocket-Building Methods

Elon Musk is making an unprecedented move by dissolving his AI startup, xAI, and folding it into his aerospace company, SpaceX, ahead of a planned public offering. This aims to package SpaceX's lucrative rocket and Starlink business with the high-cost, high-growth potential of AI. However, xAI's flagship model, Grok, has struggled to gain significant commercial or enterprise traction compared to leaders like OpenAI's ChatGPT or Anthropic's Claude. Internal turmoil led to the departure of much of xAI's founding AI talent. Musk has responded by installing SpaceX engineers as managers to transform xAI from a research lab into a high-efficiency "AI factory," focusing on infrastructure like its Colossus supercomputing cluster. Musk's vision positions the combined "SpaceXAI" as a future AI infrastructure company, addressing bottlenecks in computing power, energy, and data centers. He even proposes futuristic concepts like space-based AI data centers. To validate this story, SpaceXAI has begun sharing compute resources with former rival Anthropic. Financially, the merger appears to be a move to secure funding for xAI's massive losses by leveraging SpaceX's stable cash flow. While the combined entity targets a $1.25 trillion valuation, the market has yet to price in significant synergy. The strategic choice of SpaceX over Tesla, despite Tesla's closer ties to physical AI applications like robots and cars, is seen as Musk securing maximum control. Ultimately, Musk is betting that his proven methodology—centralized control, vertical integration, and aggressive engineering timelines—will succeed in the AI arena. But this time, he faces competitors like OpenAI and Google who are equally fast, well-funded, and determined. The merger is less about a guaranteed victory and more about ensuring Musk remains a key player at the table, regardless of the final outcome.

marsbit05/09 01:40

Dissolving xAI, Musk Wants to Rebuild an AI Company Using Rocket-Building Methods

marsbit05/09 01:40

Coinbase Q1 Earnings Report: Nearly $4 Billion Loss, Trading Volume Halved. Can AI + RWA Turn Things Around?

Coinbase's Q1 2026 earnings report revealed a net loss of $394 million, largely driven by $482 million in unrealized losses on its crypto asset holdings. Total revenue fell 31% year-over-year to $1.41 billion. Transaction revenue declined 40% to $756 million, reflecting a market-wide slump in crypto trading volumes. Consumer trading was particularly weak, down 48%, though Coinbase's global spot market share rose to a record 8.6%. Key bright spots included institutional trading revenue, which grew 37%, and a surge in derivatives activity following the Deribit acquisition. Subscription and services revenue of $584 million was more resilient, with stablecoin revenue up 11% to $305 million. Adjusted EBITDA remained positive at $303 million. Ahead of earnings, Coinbase announced a 14% workforce reduction (~700 employees) to accelerate its transition to an "AI-native" organizational model. Strategically, the company is pursuing its "Everything Exchange" vision, expanding into derivatives and predictive markets. Its partnership with Circle on USDC remains a core revenue moat, with over 25% of the stablecoin's $80 billion supply held on its platform. The company is actively engaged in shaping stablecoin legislation like the CLARITY Act. Despite significant losses and cyclical pressures, Coinbase is positioning itself as a broader on-chain financial infrastructure provider.

链捕手05/08 16:55

Coinbase Q1 Earnings Report: Nearly $4 Billion Loss, Trading Volume Halved. Can AI + RWA Turn Things Around?

链捕手05/08 16:55

活动图片