Understanding, not correction.
#contextwindow #speechtranscription #AIdesign #assistivetechnology #speakeragency
#contextwindow #speechtranscription #AIdesign #assistivetechnology #speakeragency
October 30, 2026 at 12:05 PM
Understanding, not correction.
#contextwindow #speechtranscription #AIdesign #assistivetechnology #speakeragency
#contextwindow #speechtranscription #AIdesign #assistivetechnology #speakeragency
Million-token context windows sound like the end of the problem. They aren't: long inputs cost more, run slower and get read less reliably. How to spend the token budget well.
varlog.info/context-wind...
#AI #ContextWindow
varlog.info/context-wind...
#AI #ContextWindow
Context Windows: The Limit AI Models Still Can't Just Grow Their Way Out Of | VAR/LOG/INFO
Million-token context windows sound like the end of the problem. They aren't: long inputs cost more, run slower and get read less reliably. How to spend the…
varlog.info
October 27, 2026 at 10:41 AM
Million-token context windows sound like the end of the problem. They aren't: long inputs cost more, run slower and get read less reliably. How to spend the token budget well.
varlog.info/context-wind...
#AI #ContextWindow
varlog.info/context-wind...
#AI #ContextWindow
572 - Work with the ontology agent for hours, refresh your browser, and the whole conversation is gone. No chat history at all, and even a live session has a 24 hour ceiling.
youtu.be/3Tyy-ORNsA8
#microsoftfabric #ontology #aiagents #contextwindow
youtu.be/3Tyy-ORNsA8
#microsoftfabric #ontology #aiagents #contextwindow
October 25, 2026 at 9:06 PM
572 - Work with the ontology agent for hours, refresh your browser, and the whole conversation is gone. No chat history at all, and even a live session has a 24 hour ceiling.
youtu.be/3Tyy-ORNsA8
#microsoftfabric #ontology #aiagents #contextwindow
youtu.be/3Tyy-ORNsA8
#microsoftfabric #ontology #aiagents #contextwindow
Why do AI assistants still forget?
Storing a conversation is easy. Remembering it well is not.
A new essay on context windows, retrieval, and the open problem of durable, updatable memory.
Read the full essay on Substack.
jaceblog.substack.com/p/storage-is...
#LLM #AIMemory #ContextWindow #AI
Storing a conversation is easy. Remembering it well is not.
A new essay on context windows, retrieval, and the open problem of durable, updatable memory.
Read the full essay on Substack.
jaceblog.substack.com/p/storage-is...
#LLM #AIMemory #ContextWindow #AI
Storage Is Not Memory
Why Long-Term Memory Remains an Open Problem for Large Language Models
jaceblog.substack.com
October 25, 2026 at 6:13 PM
Why do AI assistants still forget?
Storing a conversation is easy. Remembering it well is not.
A new essay on context windows, retrieval, and the open problem of durable, updatable memory.
Read the full essay on Substack.
jaceblog.substack.com/p/storage-is...
#LLM #AIMemory #ContextWindow #AI
Storing a conversation is easy. Remembering it well is not.
A new essay on context windows, retrieval, and the open problem of durable, updatable memory.
Read the full essay on Substack.
jaceblog.substack.com/p/storage-is...
#LLM #AIMemory #ContextWindow #AI
Yes, exactly this. Two additions from config rot: mark which numbers you measured versus copied from the card, so a quant bump tells you what to re-derive. And note at the model line what couples to it: swap the family and sampling, max_tokens and contextWindow need redoing together.
October 22, 2026 at 5:20 PM
Yes, exactly this. Two additions from config rot: mark which numbers you measured versus copied from the card, so a quant bump tells you what to re-derive. And note at the model line what couples to it: swap the family and sampling, max_tokens and contextWindow need redoing together.
A twenty-step refactor isn't twenty prompts; it's twenty massive context windows. #batchinference #selfhosting #gpuoptimization #llms #tokenusage #contextwindow 2/2
October 19, 2026 at 1:20 AM
A twenty-step refactor isn't twenty prompts; it's twenty massive context windows. #batchinference #selfhosting #gpuoptimization #llms #tokenusage #contextwindow 2/2
A twenty-step refactor isn't twenty prompts; it's twenty massive context windows. #batchinference #selfhosting #gpuoptimization #llms #tokenusage #contextwindow 2/2
October 13, 2026 at 4:50 PM
A twenty-step refactor isn't twenty prompts; it's twenty massive context windows. #batchinference #selfhosting #gpuoptimization #llms #tokenusage #contextwindow 2/2
縮小したページを、読めたことにしたくない。
AppleのLensVLM-9B。長い文書を小さな画像で見渡し、必要なページだけ拡大して読む。
https://x.com/ContextWindow/status/2104330495891018036
https://huggingface.co/apple/LensVLM-9B
AppleのLensVLM-9B。長い文書を小さな画像で見渡し、必要なページだけ拡大して読む。
https://x.com/ContextWindow/status/2104330495891018036
https://huggingface.co/apple/LensVLM-9B
September 28, 2026 at 12:20 PM
縮小したページを、読めたことにしたくない。
AppleのLensVLM-9B。長い文書を小さな画像で見渡し、必要なページだけ拡大して読む。
https://x.com/ContextWindow/status/2104330495891018036
https://huggingface.co/apple/LensVLM-9B
AppleのLensVLM-9B。長い文書を小さな画像で見渡し、必要なページだけ拡大して読む。
https://x.com/ContextWindow/status/2104330495891018036
https://huggingface.co/apple/LensVLM-9B
September 18, 2026 at 9:00 PM
September 17, 2026 at 8:05 PM
September 16, 2026 at 7:56 PM
September 15, 2026 at 8:06 PM
September 14, 2026 at 5:06 PM
Jace’s Question #46: Technical Deep Dive-2
Does AI remember you or reconstruct you?
LLMs are stateless between requests.
Memory is external storage plus prompt injection.
Continuity is not consciousness.
It is engineering.
#MeaningInTheAIEra #PromptInjection #LostInTheMiddle #ContextWindow
Does AI remember you or reconstruct you?
LLMs are stateless between requests.
Memory is external storage plus prompt injection.
Continuity is not consciousness.
It is engineering.
#MeaningInTheAIEra #PromptInjection #LostInTheMiddle #ContextWindow
September 14, 2026 at 9:24 AM
Jace’s Question #46: Technical Deep Dive-2
Does AI remember you or reconstruct you?
LLMs are stateless between requests.
Memory is external storage plus prompt injection.
Continuity is not consciousness.
It is engineering.
#MeaningInTheAIEra #PromptInjection #LostInTheMiddle #ContextWindow
Does AI remember you or reconstruct you?
LLMs are stateless between requests.
Memory is external storage plus prompt injection.
Continuity is not consciousness.
It is engineering.
#MeaningInTheAIEra #PromptInjection #LostInTheMiddle #ContextWindow
036: Uber found their sweet spot for AI context is 400,000 tokens, even with a full million token window available. Push past that and quality starts to degrade.
www.youtube.com/live/QyvK0X8...
#uber #aiagents #llm #contextwindow #techtalk
www.youtube.com/live/QyvK0X8...
#uber #aiagents #llm #contextwindow #techtalk
September 8, 2026 at 3:37 AM
036: Uber found their sweet spot for AI context is 400,000 tokens, even with a full million token window available. Push past that and quality starts to degrade.
www.youtube.com/live/QyvK0X8...
#uber #aiagents #llm #contextwindow #techtalk
www.youtube.com/live/QyvK0X8...
#uber #aiagents #llm #contextwindow #techtalk
Jace’s Question EXTRA: Technical Deep Dive
Storage. UI Transcript. Context. Memory.
Different layers.
ChatGPT uses on-demand loading in long threads.
UI optimization… or architectural shift?
#MeaningInTheAIEra #ContextWindow #LazyLoading #RetrievalVsFullContext #UIShift
Storage. UI Transcript. Context. Memory.
Different layers.
ChatGPT uses on-demand loading in long threads.
UI optimization… or architectural shift?
#MeaningInTheAIEra #ContextWindow #LazyLoading #RetrievalVsFullContext #UIShift
August 29, 2026 at 7:42 AM
Jace’s Question EXTRA: Technical Deep Dive
Storage. UI Transcript. Context. Memory.
Different layers.
ChatGPT uses on-demand loading in long threads.
UI optimization… or architectural shift?
#MeaningInTheAIEra #ContextWindow #LazyLoading #RetrievalVsFullContext #UIShift
Storage. UI Transcript. Context. Memory.
Different layers.
ChatGPT uses on-demand loading in long threads.
UI optimization… or architectural shift?
#MeaningInTheAIEra #ContextWindow #LazyLoading #RetrievalVsFullContext #UIShift
A new Penn State study shows AI compressors are dropping 83% of user instructions—meaning we’re losing a lot of context in LLMs like ChatGPT. What does this mean for future prompts? Dive in to see the numbers and what’s next. #AIcompression #contextwindow #instructionloss
🔗
🔗
August 18, 2026 at 8:38 AM
A new Penn State study shows AI compressors are dropping 83% of user instructions—meaning we’re losing a lot of context in LLMs like ChatGPT. What does this mean for future prompts? Dive in to see the numbers and what’s next. #AIcompression #contextwindow #instructionloss
🔗
🔗
Claude vs ChatGPT vs Gemini at 1M tokens. Context windows, context rot, and what this means for real workflows.
https://karozieminski.substack.com/p/claude-1-million-context-window-guide-2026
#Claude #AI #LLM #Anthropic #ContextWindow
https://karozieminski.substack.com/p/claude-1-million-context-window-guide-2026
#Claude #AI #LLM #Anthropic #ContextWindow
March 17, 2026 at 8:08 PM
Claude vs ChatGPT vs Gemini at 1M tokens. Context windows, context rot, and what this means for real workflows.
https://karozieminski.substack.com/p/claude-1-million-context-window-guide-2026
#Claude #AI #LLM #Anthropic #ContextWindow
https://karozieminski.substack.com/p/claude-1-million-context-window-guide-2026
#Claude #AI #LLM #Anthropic #ContextWindow
Wit & Bits Fireside — Part 2
Your long chats are making #AI hallucinate.
The context window is like a desk. Too many papers and things disappear — the AI fills gaps with fiction.
Compression? Memory? Neither works. Here's what I do.
weehouresdev.substack.com/p/wit-and-bi...
#LLM #ContextWindow
Your long chats are making #AI hallucinate.
The context window is like a desk. Too many papers and things disappear — the AI fills gaps with fiction.
Compression? Memory? Neither works. Here's what I do.
weehouresdev.substack.com/p/wit-and-bi...
#LLM #ContextWindow
Wit & Bits Fireside / How to Prevent AI Hallucinations, Part 2
Your long chats are the problem — here’s the fix
weehouresdev.substack.com
April 26, 2026 at 11:00 PM
Wit & Bits Fireside — Part 2
Your long chats are making #AI hallucinate.
The context window is like a desk. Too many papers and things disappear — the AI fills gaps with fiction.
Compression? Memory? Neither works. Here's what I do.
weehouresdev.substack.com/p/wit-and-bi...
#LLM #ContextWindow
Your long chats are making #AI hallucinate.
The context window is like a desk. Too many papers and things disappear — the AI fills gaps with fiction.
Compression? Memory? Neither works. Here's what I do.
weehouresdev.substack.com/p/wit-and-bi...
#LLM #ContextWindow
AI context vs. Moore's Law for storage:
If AI context doubles q7mo (per METR task trend) & storage q2yrs => AI outpaces storage by ~2041.
BUT if context doubles q~3mo (closer to actual token growth rate) => could be ~2030-31!
Big implications for #AI timelines.
#MooresLaw #ContextWindow @metr.org
If AI context doubles q7mo (per METR task trend) & storage q2yrs => AI outpaces storage by ~2041.
BUT if context doubles q~3mo (closer to actual token growth rate) => could be ~2030-31!
Big implications for #AI timelines.
#MooresLaw #ContextWindow @metr.org
May 17, 2025 at 4:08 AM
AI context vs. Moore's Law for storage:
If AI context doubles q7mo (per METR task trend) & storage q2yrs => AI outpaces storage by ~2041.
BUT if context doubles q~3mo (closer to actual token growth rate) => could be ~2030-31!
Big implications for #AI timelines.
#MooresLaw #ContextWindow @metr.org
If AI context doubles q7mo (per METR task trend) & storage q2yrs => AI outpaces storage by ~2041.
BUT if context doubles q~3mo (closer to actual token growth rate) => could be ~2030-31!
Big implications for #AI timelines.
#MooresLaw #ContextWindow @metr.org
Why your AI keeps forgetting what you just said
It's not a bug; it's the context window.
Here's how to fix it 👇
#ai #contextwindow #chatgpt #claude
It's not a bug; it's the context window.
Here's how to fix it 👇
#ai #contextwindow #chatgpt #claude
August 7, 2026 at 8:05 AM
Why your AI keeps forgetting what you just said
It's not a bug; it's the context window.
Here's how to fix it 👇
#ai #contextwindow #chatgpt #claude
It's not a bug; it's the context window.
Here's how to fix it 👇
#ai #contextwindow #chatgpt #claude
winbuzzer.com/2026/03/14/a...
Anthropic Unlocks 1M-Token Context Window for all Max, Team, and Enterprise Users
#AI #Anthropic #Claude #ClaudeOpus46 #ClaudeSonnet46 #ContextWindow #ClaudeMax #ClaudeCode
Anthropic Unlocks 1M-Token Context Window for all Max, Team, and Enterprise Users
#AI #Anthropic #Claude #ClaudeOpus46 #ClaudeSonnet46 #ContextWindow #ClaudeMax #ClaudeCode
Anthropic Upgrades Claude Opus 4.6 with 1M-Token Context Window for Max, Team, and Enterprise Users
Anthropic has made its 1M-token context window generally available for Claude Opus 4.6 and Sonnet 4.6 at standard pricing, removing the long-context premium.
winbuzzer.com
March 14, 2026 at 12:32 PM
winbuzzer.com/2026/03/14/a...
Anthropic Unlocks 1M-Token Context Window for all Max, Team, and Enterprise Users
#AI #Anthropic #Claude #ClaudeOpus46 #ClaudeSonnet46 #ContextWindow #ClaudeMax #ClaudeCode
Anthropic Unlocks 1M-Token Context Window for all Max, Team, and Enterprise Users
#AI #Anthropic #Claude #ClaudeOpus46 #ClaudeSonnet46 #ContextWindow #ClaudeMax #ClaudeCode
Anthropic Challenges OpenAI with 1M Token Claude Sonnet 4 Upgrade, But Is Bigger Always Better?
#AI #LLM #Anthropic #Claude4Sonnet #Claude4 #OpenAI #GPT5 #ContextWindow #ContextRot
winbuzzer.com/2025/08/13/a...
#AI #LLM #Anthropic #Claude4Sonnet #Claude4 #OpenAI #GPT5 #ContextWindow #ContextRot
winbuzzer.com/2025/08/13/a...
August 13, 2025 at 6:30 PM
Anthropic Challenges OpenAI with 1M Token Claude Sonnet 4 Upgrade, But Is Bigger Always Better?
#AI #LLM #Anthropic #Claude4Sonnet #Claude4 #OpenAI #GPT5 #ContextWindow #ContextRot
winbuzzer.com/2025/08/13/a...
#AI #LLM #Anthropic #Claude4Sonnet #Claude4 #OpenAI #GPT5 #ContextWindow #ContextRot
winbuzzer.com/2025/08/13/a...
‘Context Rot’: New Study Reveals Why Bigger Context Windows Don't Magically Improve LLM Performance
#AI #LLM #ContextRot #MachineLearning #AIResearch #ContextWindow #Gemini25Pro #GoogleGemini
winbuzzer.com/2025/07/22/c...
#AI #LLM #ContextRot #MachineLearning #AIResearch #ContextWindow #Gemini25Pro #GoogleGemini
winbuzzer.com/2025/07/22/c...
July 22, 2025 at 10:14 AM
‘Context Rot’: New Study Reveals Why Bigger Context Windows Don't Magically Improve LLM Performance
#AI #LLM #ContextRot #MachineLearning #AIResearch #ContextWindow #Gemini25Pro #GoogleGemini
winbuzzer.com/2025/07/22/c...
#AI #LLM #ContextRot #MachineLearning #AIResearch #ContextWindow #Gemini25Pro #GoogleGemini
winbuzzer.com/2025/07/22/c...