DeepSeek
DeepSeek: Frontier model, AI assistant, or foundation-model company shaping general AI interfaces and next-generation human-computer interaction.
ABAB Structured Brief
DeepSeek is indexed in ABAB Crypto Map under AI Models & Apps. This page keeps the official site, category, tags, and related ABAB coverage together as a searchable crypto project profile. Official domain: deepseek.com.
Related News & Analysis
DeepSeek V4 Pro API Suspected of Routing to Claude Fable 5
Leaker blogger Leo questions whether the official API of DeepSeek V4 Pro routes complex programming requests to Claude Fable 5 for collecting outputs for model distillation. Testing during the generation of 3D g...
From OpenSea to OpenRouter: Alex Atala Analyzes Multi-Model Paradigms, Jevons Paradox, and Dynamic Cost Control for Enterprises
"OpenRouter CEO: Why Chinese Open Models Are Beating the US Why Enterprises Fear OpenAI & Anthropic" (20VC interview with Harry Stebbings, featuring OpenRouter co-founder and CEO Alex Atala), here are the key points summarized: 1. From OpenSea to OpenRouter: High-Concurrency Architecture and Market Evolution • Lessons from OpenSea: Alex was a co-founder of the NFT trading platform OpenSea. Early on, OpenSea experienced massive traffic surges and server downtime risks. He brought the underlying architecture experience of high concurrency, high availability (Uptime), and elastic scaling to OpenRouter, ensuring stability during model surges or service fluctuations. • Rise of Inference Providers: It was initially thought that model hosting would be monopolized by the three major cloud providers (AWS, Azure, GCP), but in reality, specialized inference providers like Fireworks and Together respond faster and perform better in deploying open-weight models (such as GLM, Kimi, DeepSeek). • Nvidia's ecosystem preference: Nvidia tends to diversify customer concentration by allocating GPU quotas to multiple inference providers, fostering a flourishing ecosystem of underlying computing power providers. 2. Multi-Model Future and AI Neurodiversity • Rejecting single-model monopoly: Advocating for "AI Neurodiversity," firmly believing that the future will not be dominated by a single model. Both enterprises and individuals need to use a combination of different models to achieve higher creativity and cost-effectiveness. • Specialization and brand intelligence: Enterprises will not rely solely on a generic model in the future but will fine-tune or train proprietary models (such as using LoRA plugins) for their core business while also utilizing other excellent open-source/closed-source models across the network. • Jevons Paradox validation: Taking GPT-5.6 / Luna as an example, after OpenAI reduced its price by 10 times, usage on the OpenRouter platform surged by 13 times. Lowering model prices does not reduce total expenditure; instead, it exponentially stimulates a larger demand for calls. 3. Why Enterprises Remain Cautious of Closed-Source Giants like OpenAI & Anthropic • Preventing vertical encroachment by giants (e.g., Claude Design vs. Figma): Model vendors have strong incentives to enter vertical application scenarios (e.g., Anthropic launching Claude Design). Enterprises worry that direct ties to closed-source giants will lead to opaque data policies, binding risks, and potential vertical replacement by the giants. • Data risks and VPC needs: Many enterprises find it difficult to fully trust closed-source vendors' data retention and privacy policies, preferring to deploy open-weight models in their own VPC (Virtual Private Cloud) or through open gateways for greater control. 4. The Competition of Open-Source Models Between China and the US: The US is Lagging • Strong momentum of Chinese open-source models: In the open-weight domain, Chinese open-source models (such as DeepSeek, GLM 5.2, Kimi/Moonshot, Qwen, etc.) have made significant breakthroughs in performance, inference efficiency, and writing capabilities. The US is currently lagging in the open-source model field. • Developer usage preferences: In the OpenRouter's ranking of open-source/open-weight model usage, Chinese open-source models have long occupied the top positions. • Distillation and catch-up strategies: Distillation is a conventional scientific method to enhance model efficiency. US Neolabs (new large model laboratories, such as Poolside, Thinking Machines) can quickly catch up through compliant distillation and reinforcement learning (RL), provided they solve the barriers to acquiring computing power. 5. Harness, Agent Architecture, and New Paradigms in Enterprise Management • Difference between Harness and Apps: Harness is built on Unix/command line principles as an Agent control layer, which is more composable, deterministic, and model-friendly than traditional API or UI-based Apps. • Orchestrator and Sub-Agent architecture: The mainstream architecture of the future will be a high-IQ "main orchestration model" coordinating the overall situation, issuing instructions to multiple low-cost, high-deterministic "open-source sub-agents" to execute standardized tasks such as classification and extraction. • Dynamic Employee Cost: Enterprise management will undergo transformation in the AI era. The inference costs incurred by employees using different models are highly dynamic, and in the future, enterprises will need to manage performance and costs based on the match between "employee output" and "AI computing power consumption costs."
OpenMed Claims Open Source Cutting-Edge AI Surpasses Closed Source Models This Week, Local Deployment is the Bottom Line for Medical Data Sovereignty
...0b, Nemotron 3.5 Lightning, Grok Bot, Qwen series, GLM-5.3, DeepSeek v4 Pro, etc. It further added that "faster treatment for patients, faster cures for all diseases" is its core message. OpenMed focuses on o...
Exclusive Interview with NVIDIA CEO Jensen Huang: Debunking AI Doomsday Theories, Optimistic About the Future of Open Models and Computing Infrastructure in the Next Decade
This episode of "Jensen Huang says the AI doomers have it wrong" (Axios interview, conducted by Mike Allen) summarizes the core content as follows: 1. Geopolitics, Export Controls, and Open Source Ecosystem • Views on US-China geopolitical competition: hopes for open technology and research exchanges. Half of the world's AI researchers are in China, contributing groundbreaking research. • Core value of open models: • The market needs not only closed-source services like Anthropic and OpenAI but also open models. • Openness is the foundation of scientific progress, cybersecurity, national security, and economic security. Through open and transparent oversight and testing, the community can better strengthen security defenses (cybersecurity requires distributed self-defense). • Use of Chinese open models (such as Kimi, DeepSeek, Qwen, etc.): • US companies should be allowed to use excellent Chinese open models, with no so-called "backdoor" risks. Models typically run in isolated and secure "sandboxes" and "harnesses." • Emphasizes that excellent open models will promote the entire industry's use of AI, and increased usage means a greater need for NVIDIA's computing infrastructure and data centers, which is a significant benefit for the entire industry and hardware suppliers. 2. Refuting "AI Doomsday" and "Job Loss Panic" • Criticism of "AI doomsday theories": • Believes that some tech leaders' claims that "AI will destroy humanity" or "will eliminate half of US jobs" are baseless fantasies and hype. • This fictional doomsday narrative only creates unnecessary social panic and hinders the rapid application of AI in the industry. • AI creates jobs rather than destroys them: • Historical and current facts prove that productivity improvements lead to more industry opportunities. • Automation increases efficiency (e.g., radiologists automatically reviewing scans, paralegals automatically processing documents), significantly enhancing service capacity and increasing the number of patients/customers, which in turn drives further demand for these positions. • The current AI boom has directly driven an explosion of jobs in manufacturing, energy, chips, and infrastructure sectors. • The distinction between jobs and tasks: AI automates specific "tasks," not the ultimate "purpose" of a profession. For example, typing is just a task; solving problems and creating value are the essence of work itself. 3. NVIDIA's Product Layout and Market Forecast • NVIDIA Nemotron open model: not aimed at competing with closed-source giants in the general large model space, but rather providing customizable open foundational models for enterprises needing sovereignty, privacy protection, and domain-specific knowledge. • Computing expenditure and "bubble theory": • Denies the danger of "AI computing surplus/bubble" in the short term. AI is currently in the earliest stage of infrastructure development (unlikely to burst in the next 5-10 years). • The current bottleneck lies in physical constraints (chip, memory, power, land, and a shortage of data center construction workers), which can actually prevent overheating. • Long-cycle transformation of the semiconductor industry: • This round of prosperity is not driven by cyclical or seasonal consumer demand but by a historic reconstruction of computer infrastructure (from pre-recorded static content to real-time generated intelligent tokens). • Predicts that the semiconductor industry needs to expand 5 to 10 times in scale within the next 10 years to meet demand. 4. Evolution of AI and the Era of Robotics • The essence and value of tokens: Tokens are mathematical units embedded with knowledge and intelligence. As models become smarter, the value contained in each token increases, leading customers to be more willing to pay, forming a commercial flywheel. • The era of agentic AI: In the future, there will be hundreds of billions or even trillions of AI agents running in the background and collaborating globally, leading to explosive demand for computing infrastructure geometrically. • The "ChatGPT moment" for robots: The "ChatGPT moment" for robots has arrived (showing surprising reasoning and spatial understanding abilities, such as autonomously understanding to open the cabinet door before placing the apple); it is expected that within the next 3 to 4 years, robots will truly enter a stage of widespread and practical use. 5. Leadership and Personal Philosophy • Pain and tempering: Excellence comes from "a lot of pain and suffering" and repeated practice when no one is watching. These setbacks shape character, resilience, and a calm mindset when facing high pressure. • Confidence in America: Still believes in the "American Dream." The US has an open discussion environment, a soil for free innovation and collaboration, and as long as it continues to embrace and apply new technologies, it can maintain its lead in the next industrial revolution. • Time management philosophy: Does not wear a watch or strictly constrain himself with Outlook, maintaining a focus on "the present moment (Now is the most important time)" and being highly focused on spending time with family and pets on weekends.
Li Bojie Complains on Social Media About Cheating Allegations During DeepSeek Coding Interview, Interview Terminated on the Spot
... on social media that he was accused of cheating during the DeepSeek coding interview, which led to the immediate termination of the interview; this incident quickly prompted former investor Du Jun (co-founder of ABCDE) ...
DeepSeek and Alibaba Financing Negotiations Break Down
DeepSeek's negotiations for a massive financing round with Alibaba have ultimately broken down. The core disagreement between the two parties lies in Alibaba's desire to deeply integrate its AI ecosystem (including Tongy...