Glossary
FAQ
Search
Frequently Asked Questions
Common questions about AI, models, and tools.
AI Concepts
Can sub-quadratic architectures replace attention?
Can sub-quadratic models replace transformers?
What are hybrid attention architectures?
What is self-attention in transformers?
What is the difference between encoder and decoder attention?
What is the quadratic cost problem with attention?
Why did transformers replace RNNs?
Why is attention O(n²) and why does it matter?
Why is the √d scaling factor important in attention?
AI Development Tools
Are LLMs better at reviewing code or writing it?
Are multi-agent AI workflows more accurate than single agents?
Can Claude Code spawn multiple agents to work on tasks?
How does Xcode's Claude integration differ from GitHub Copilot or Cursor?
How much do multi-agent workflows cost compared to single agents?
What are agent teams in Claude Code?
What is multi-agent orchestration in Claude Code?
What is the Claude Agent SDK integration in Xcode 26.3?
When is Xcode 26.3 with Claude Agent SDK available?
AI Infrastructure
Does the GPT-5.2 speed improvement apply to ChatGPT?
How did OpenAI achieve the 40% inference speedup?
Is 'price per million tokens' comparable across LLM providers?
Is the GPT-5.2 speed improvement a real 40% speedup or just shorter outputs?
What is Byte Pair Encoding and how do LLMs use it?
What is the difference between prefill and decode in LLM inference?
What is the KV cache and why does it matter for inference cost?
Why are LLM API costs higher for non-English languages?
Why can't ChatGPT count the Rs in 'strawberry'?
Why do longer context windows cost more for LLM inference?
Why is LLM inference so much less efficient than training?
AI Models
How does Claude Opus 4.6 compare to GPT-5.2?
How much does Claude Opus 4.6 cost?
What is Claude Opus 4.6?
AI Policy
How does Trusted Access for Cyber differ from OpenAI's existing safety measures?
What is OpenAI's Trusted Access for Cyber program?
Why is OpenAI requiring identity verification for cybersecurity tools?
Will other AI labs follow OpenAI's tiered access approach?
AI Research
Can Claude AI actually run experiments in a biology lab?
Can I train GPT-2 on cloud spot instances without interruption?
Can other labs replicate OpenAI and Ginkgo's AI-driven protein synthesis results?
Can other research labs use AI to optimize their experiments like this?
Do GPT-2 training optimizations scale to larger models?
How did GPT-5 achieve the 40% cost reduction in protein synthesis?
How did GPT-5 optimize protein synthesis costs by 40%?
How does Anthropic use the constitution for training?
How does the Allen Institute AI partnership differ from the HHMI collaboration?
How much did GPT-5 reduce protein synthesis costs?
How much does it cost to train GPT-2 in 2025?
What are Anthropic's new scientific research partnerships?
What is Claude's new constitution?
What is nanochat?
What is RLHF and how does it work?
What is the alignment tax in AI models?
What is the difference between DPO and RLHF?
What optimizations made GPT-2 training 600× cheaper?
Why did Anthropic shift from rules to reasoning in Claude's constitution?
Why does Anthropic emphasize interpretability in its science AI partnerships?
Why is GPT-2 being called 'the new MNIST'?
Why is RLHF alignment so easy to break?
AI Safety
Why is OpenAI gating cybersecurity capabilities behind identity checks?
Biotech
What is cell-free protein synthesis?
Business
How much does the Google-Apple AI deal cost?
How much is Alphabet spending on AI infrastructure in 2026?
Is OpenAI adding ads to ChatGPT?
What is the Google-Apple AI deal?
What is the Streisand Effect in the context of Altman's response?
What was Anthropic's Super Bowl ad about?
Why did Sam Altman respond to Anthropic's anti-ad campaign?
Why did Sundar Pichai avoid the Apple question on the earnings call?
Why did Sundar Pichai ignore the analyst question about Apple on Alphabet's earnings call?
Will OpenAI introduce ads to ChatGPT?
Enterprise AI
How does OpenAI Frontier differ from Anthropic's approach to enterprise AI agents?
How does OpenAI Frontier differ from Anthropic's Claude Code?
What enterprise systems does Frontier connect to?
What is OpenAI Frontier?
What problems does Frontier solve for enterprises?
Industry Strategy
Why are AI companies investing in education?
Partnerships
What is Anthropic's partnership with Teach For All?
Policy & Business
Did Anthropic promise Claude will never have ads?
How does Anthropic plan to compete with Microsoft in India?
How does Anthropic plan to make money without ads?
How will Anthropic compete with Microsoft Copilot in India?
How will Anthropic fund Claude without advertising revenue?
What is the ratchet effect in AI advertising?
Who is Irina Ghose and why did Anthropic hire her?
Why does Anthropic say AI advertising is different from search advertising?
Why is Anthropic pledging that Claude will never run ads?
Why is India important for Anthropic and Claude?
Policy & Regulation
Can the UK government independently maintain the Claude-powered system?
What is Anthropic building for GOV.UK?
What is Anthropic's Long-Term Benefit Trust?
What powers does Anthropic's Trust actually have?
Who are the current trustees of Anthropic's Long-Term Benefit Trust?
Who is accountable if the GOV.UK AI assistant gives wrong advice?
Who is Tino Cuéllar and why does his appointment matter?
Why is starting with employment support risky for a government AI deployment?