Bilingual

Because It Speaks in Words
6 days ago
- Understanding goes beyond mere knowledge, requiring a moment of deep connection with truths.
- The moon, Jupiter, and Venus symbolize a shared celestial view across history, highlighting our limited human perspective.
- Words and stories are powerful tools for communication and understanding, rooted in our innate desire for narrative.
- Stories help us make sense of the world, becoming part of our identity and shaping our perceived truths.
- Disinformation and misinformation are rampant today, amplified by chaotic and numerous voices in modern society.
- Large Language Models (LLMs) have gained mainstream attention because they communicate in words, unlike other AI techniques that use vectors or numbers.
- LLMs mimic human speech without true understanding, leading us to trust their plausible but mindless output.
- The human need for stories is being exploited by technology, similar to how processed food hijacks biological cravings.
- LLMs may push us toward statistical averages, but as AI-generated text diverges from objective meaning, stability is uncertain.
- Ultimately, only humans can truly tell and learn from stories, with the quest for truth remaining a lonely human endeavor amid seductive digital voices.
Anthropic says US lifts export ban on Fable 5
5 days ago
- US government lifts export ban on Anthropic's advanced AI models Claude Fable 5 and Mythos 5.
- The ban was initially imposed on June 12 due to national security concerns about potential misuse by hackers.
- Anthropic addressed risks by agreeing to detect security issues, collaborate on future releases, and alert the government of malicious activity.
- Models were restored on Wednesday, with the Commerce Department reserving the right to reconsider the decision if necessary.
- Fable 5 is for consumer use, while Mythos 5 targets businesses and cybersecurity experts, both released on June 9.
- The company disagreed with the initial suspension, arguing that a potential jailbreak finding shouldn't warrant recalling widely deployed models.
Why Everyone Is Suddenly Talking About 'Universal Basic Capital'
3 days ago
- Universal Basic Capital (UBC) is proposed as a hedge against AI-induced inequality by giving everyone an ownership stake in AI companies.
- UBC differs from Universal Basic Income (UBI) by providing equity accounts instead of cash payments, potentially avoiding political hurdles related to taxes and welfare.
- The policy has bipartisan appeal, with examples like the 'Trump Accounts' bill and proposals from figures like Vivek Ramaswamy to grant stock accounts at birth.
- Critics warn that a government-controlled sovereign wealth fund model, as proposed by Bernie Sanders, could lead to regulatory capture and politicization of AI companies.
- Implementing UBC without careful design risks giving excessive control to political actors, undermining corporate independence and public oversight.
Data Science Weekly – Issue 658
3 days ago
- Data Science Weekly newsletter #659 from June 26, 2026, curated by Hannah & Sebastian, covers Data Science, AI, ML, Data Visualization, and ML/Data Engineering topics with editor's picks, articles, and community polls.
- Editor's picks include an article on colors beyond digital display, using Bradley-Terry model for dog treat preferences, and generative AI creating optimized burgers for taste, sustainability, and nutrition.
- Notable articles discuss open-source ATS scoring resumes variably, underrated data science skills, AI's role in data science, vibe coding vs. engineering, regression coefficient issues, and tools like Rust tensor stacks, Typst graphics, geo-experiment benchmarks, and CUDA kernel processes.
Essential Until It Wasn't
2 days ago
- The nature of 'programming' has evolved over time, with each era's practitioners believing their skills to be the permanent core of the craft, such as memorizing opcodes or managing memory manually.
- Fred Brooks' distinction between essential complexity (inherent to the problem) and accidental complexity (tool-related friction) is useful, but the boundary between them shifts as tools absorb tasks, making what was once essential become accidental.
- Historical examples show this boundary shift: opcode memorization, manual memory management, and API knowledge have all transitioned from essential to accidental as new tools emerged, often met with gatekeeping complaints like 'they're not real programmers.'
- AI is now moving the boundary past implementation-level reasoning, making tasks like writing code character-by-character and manual debugging accidental, while the essential work shifts to specification, architecture, and theory-building.
- This transition unbundles traits: implementation-specific traits (e.g., patience with syntax) become less relevant, while architecture-level traits (e.g., systems thinking, curiosity) persist, causing disorientation for practitioners adapted to the old boundary.
- The essay argues that programming's essence lies in theory-building and conceptual integrity, not code-writing, and that the current boundary shift is not final—future shifts may affect even architecture or specification.
- A recipe for understanding such transitions involves identifying fixed distinctions, exploring historical precedents, listening to gatekeeping complaints, and determining what remains essential after the boundary moves.
China's ByteDance discovers new scaling law that could sustain AI boom
2 days ago
- ByteDance's Seed AI team found that AI agents can double their learning speed every three months through interaction with real-world environments.
- The AI industry is shifting from brute-force data and compute scaling to agentic AI, but post-deployment learning in real-world settings is poorly understood.
- To study this, ByteDance created EdgeBench, a benchmark with 134 long-term tasks requiring over 12 hours of continuous AI agent operation across various domains.
LLM's as a Different Kind of Intelligence
14 hours ago
- LLMs represent a different kind of intelligence, distinct from human intelligence.
- Human analogies like 'junior engineer' or 'abstraction layer' are flawed because LLMs lack intent and deterministic reasoning.
- LLMs combine vast knowledge with an inability to understand intent, leading to subtle drift from user goals over time.
- Probabilistic behavior from LLMs challenges the deterministic foundations of software engineering.
- Effective use requires learning LLMs' unique failure modes rather than forcing familiar comparisons.
The Electricity: Why Google may be trying to make intelligence disappear
16 hours ago
- Google's foundational AI paper 'Attention Is All You Need' introduced the transformer architecture that underpins modern AI models like ChatGPT.
- Despite nearly being disrupted by its own technology, Google continues to openly share AI research and models, such as Gemma, rather than locking them down.
- Google's strategy is not to win the 'smartest model' race but to integrate AI as ubiquitous infrastructure, like electricity, across its products and platforms.
- By embedding AI (e.g., Gemini Nano) on-device in Android, Google shifts compute costs to users' devices, reducing its own operational expenses.
- Google dominates key digital access points: Search, Android, Chrome, and Workspace, ensuring AI touches users through existing Google-controlled channels.
- Google is investing in physical infrastructure, like power companies and grid projects, to support its data centers and long-term AI energy needs.
- A potential risk to Google's strategy is the rising cost of memory due to AI demand, which could increase phone prices and affect its on-device AI plans.
- The goal is for AI to become invisible—users won't recall specific models, as AI seamlessly integrates into daily tools and environments.

Hasty Briefsbeta

#artificial intelligence

Because It Speaks in Words

Anthropic says US lifts export ban on Fable 5

Why Everyone Is Suddenly Talking About 'Universal Basic Capital'

Data Science Weekly – Issue 658

Essential Until It Wasn't

China's ByteDance discovers new scaling law that could sustain AI boom

LLM's as a Different Kind of Intelligence

The Electricity: Why Google may be trying to make intelligence disappear