Issue 11 · one issue a week — the judgment is continuous
JustSayAIWeeklyReport
Greenberg Traurig Deploys Agentic AI for Legal Work, Bill Gates Urges AI Limits, Google Upgrades Long-Context Agents
人民公园说AI · JustSayAI
www.justsayai.org
Week of
08.25 — 08.31
2026-08-31 · the week's read
Contents · Chapters
This issue,
what to read.
The timeline on the left mirrors these chapters — read in order, or jump straight to any story.
Editor’s Thesis
The Week's Judgment
Greenberg Traurig Deploys Agentic AI for Legal Work,Bill Gates Urges AI Limits,
Google Upgrades Long-Context Agents
As legal AI and long-context optimization techniques push Agents into specialized deep waters, Gates' restrictive order and the turbulence within Google's safety team serve as reminders that leaps in capability are accompanied by loosening safety guardrails; notably, the resonance between external regulatory calls and internal talent reassignments suggests that if the industry fails to establish constraints at this juncture, professional penetration could instead become an accelerator for risk amplification.
The Tape · This Week's Audio
The full 07:41. Hear the first 60s free.
JustSayAI Podcast · Weekly issue 11 audio
First 60s free · full 07:41 for subscribers
Key Moments · every red tick is a marked moment in this week's audio
00:23in preview
Greenberg Traurig Rolls Out Agentic CoCounsel Legal AI Across Global Offices
“Greenberg Traurig rolled out Thomson Reuters' CoCounsel Legal across all their global offices.”
- Greenberg Traurig deploys Thomson Reuters CoCounsel Legal across global offices, marking among first global-scale Agent AI implementations in legal industry.
- AI serves as assistant, not decision-maker, handling document review, due diligence, contract comparison, with human oversight required for critical outputs
- Tech adoption deepens law firm divide: big firms use AI to cut hours and boost margins, small firms face liability gaps.
02:04locked · past preview
Bill Gates Warns: Unfettered AI Will Do More Harm Than Good
“The same week they were about to drag AI into the courtroom, Gates dumped a bucket of ice water over everyone's heads.”
- Bill Gates warns AI's harm may outweigh benefits without limits, ponders explaining to future generations.
- He proposed taxing AI/robots, reserving human-only accountability roles like judges, surgeons, and airline captains.
- Industry discussion shifted from "whether to restrict" to "how to restrict," marking regulatory progress.
03:26locked · past preview
Google Slashes Long-Context Token Costs by 16x with State Tracking
“The most hardcore move is actually Google's—slashing long-session token usage by 94%.”
- Google shifts AI long chats from "full recording" to "live note-taking," logging only key state changes, cutting token usage by 94%.
- Costs slashed to 1/16, long-context Agent apps may shift from burning demos to deployable products.
- Explicit state definitions risk cascading errors from misrecorded conclusions, with designer accountability.
Market Tape · The Week's Prices
Prices first, judgment after.
3 AI names · 1 up · 2 downBest 华宇软件 +7.16%Worst 汤森路透 -2.20%2026-08-25 → 2026-08-31 · Closing prices against the last session before the window. Those are facts; what we draw from them is judgment.
华宇软件300271.SZ · week's best
+7.16%
5.31 → 5.69
谷歌-AGOOGL
-0.42%
348.06 → 346.59
worst session -1.43%（2026-08-26）
worst session -1.43%（2026-08-26）
汤森路透TRI · week's worst
-2.20%
108.62 → 106.23
worst session -3.90%（2026-08-25）
worst session -3.90%（2026-08-25）
UpDownWeekly change · source: third-party market data
Shanghai Composite -0.11%（3952.18）Nasdaq Composite -0.52%（26402.42）
Story 01 · Tape 00:23 · AI Industry News
Greenberg Traurig Rolls Out Agentic CoCounsel Legal AI Across Global Offices
Greenberg Traurig rolled out Thomson Reuters' CoCounsel Legal across all their global offices.
Fact
🤖 Greenberg Traurig Rolls Out CoCounsel Legal AI Globally
- 01Greenberg Traurig deploys Thomson Reuters CoCounsel Legal across global offices, marking among first global-scale Agent AI implementations in legal industry.
- 02AI serves as assistant, not decision-maker, handling document review, due diligence, contract comparison, with human oversight required for critical outputs
- 03Tech adoption deepens law firm divide: big firms use AI to cut hours and boost margins, small firms face liability gaps.
Source ↗ unite.ai
Story 02 · Tape 02:04 · Technology
Bill Gates Warns: Unfettered AI Will Do More Harm Than Good
The same week they were about to drag AI into the courtroom, Gates dumped a bucket of ice water over everyone's heads.
Fact
🤖 Bill Gates Urges Major AI Restrictions, or Harms Will Outweigh Benefits
- 01Bill Gates warns AI's harm may outweigh benefits without limits, ponders explaining to future generations.
- 02He proposed taxing AI/robots, reserving human-only accountability roles like judges, surgeons, and airline captains.
- 03Industry discussion shifted from "whether to restrict" to "how to restrict," marking regulatory progress.
Source ↗ cnn.com
Story 03 · Tape 03:26 · Others
Google Slashes Long-Context Token Costs by 16x with State Tracking
The most hardcore move is actually Google's—slashing long-session token usage by 94%.
Fact
🤖 Google Cuts Long-Context Agent Token Usage by 94% with New State-Tracking Method
- 01Google shifts AI long chats from "full recording" to "live note-taking," logging only key state changes, cutting token usage by 94%.
- 02Costs slashed to 1/16, long-context Agent apps may shift from burning demos to deployable products.
- 03Explicit state definitions risk cascading errors from misrecorded conclusions, with designer accountability.
Source ↗ reddit.com
Signal Scan · The Week's Other Signals
3 more that missed the main thread.
🤖 Google Removes 90 AI Safety Experts from DeepMind, Sparking Internal Concerns
This issue05🤖 Hugging Face Open-Sources 25cm Bipedal Robot Microduck, $399
This issue06🤖 Google Offers Another Perk, Eligible Users Can Claim
This issueThe full week — every daily brief's headline, linked to its issue:
08.25Alabama Subpoenas OpenAI Over Breach, Students' AI Deepfakes Target Teachers With No Accountability - AI Daily BriefOpenAI faces Alabama probe over model safety flaws; student AI abuse sparks liability crisis, while ToMoE dynamic pruning and JetBrains local deployment push efficiency and privacy forward.
AM08.25AI Hits Entry Jobs; Data Centers' Killer App: Solid-State TransformersStanford: AI hits entry-level jobs hardest; next-gen models: ontological shock.
PM08.26Figure.AI Drops Index Robot Dataset—AI Job Applications Now Killing Hiring Speed | AI Daily BriefFigure.AI drops world's largest robot dataset; DeepSeek's cheap models hit record US share; Perplexity & GEEKOM race into on-device AI—signaling compute is escaping the cloud and invading physical reality, even as hiring freezes bite.
AM08.26Most "Agentic AI" Is Just Workflow in Disguise, Humanoid Robot Olympics Break Records and Catch Fire | AI Daily BriefAgentic AI exposed as glorified flowcharts, yet IBM Granite 4.2 doubles down on enterprise Agents; humanoid robots set records and catch fire; OpenAI exec exodus and Google hype wars reveal sector's bubble pains.
PM08.27JustSayAI AI Daily Brief (08/27): Gates Urges AI Limits, Models FailGates urges AI limits: harm>benefit; Salesforce CRM + Claude; LLMs flop
AM08.27Nvidia Doubles Profit to $59.7B, Amazon Chip Orders 3x: AI Daily BriefNVIDIA profit doubles to $59.69B on AI spending
PM08.28NVIDIA Buys HuggingFace for $13B; Waymo/Zoox Report 24+ Injuries in 2YNvidia's $13B Hugging Face buy: open-source AI; tariffs, injuries, brain drain.
AM08.28Tencent Open-Sources Hy4-preview MoE, xAI Sued for Child Abuse Content in Grok Training | AI Daily BriefTencent drops 770B Hy4-preview LLM, igniting open-source race; xAI hit with class-action over training data; ex-OpenAI Barret Zoph bolts back to Google.
PM08.29AI Daily Brief: Anthropic Auto-Align AI > Humans, Claude Agents Fix 10Anthropic auto-align>humans; Claude fixes 10 fails; giants join Agent-AI defense.
AM08.29Hugging Face Unveils Open-Source Humanoid Robot, Lambda Raises $1B for AI Compute Leasing Microsoft Infrastructure - AI Daily BriefHugging Face drops $399 Microduck: 25cm open-source bipedal bot; Lambda secures $1B debt to buy NVIDIA chips for Microsoft; US cities rush to cancel Flock surveillance contracts.
PM08.30Sony, Warner Sue Anthropic for Billions: JustSayAI AI Daily Brief 8/30Sony/Warner sue Anthropic for lyrics; Nvidia buys Hugging Face $12.9B; Fed on AI
AM08.30Nvidia's $12.9B Hugging Face Buy: NIM Error - JustSayAI AI Daily BriefNVIDIA acquires Hugging Face for $12.9B; Anthropic unveils Agent safety standard.
PMThis week ran 6 headlines; 3 made the main thread; 12 daily briefs.
The Archive · Past Issues
One issue a week. The judgment is continuous.
2026.08.31Greenberg Traurig Deploys Agentic AI for Legal Work, Bill Gates Urges AI Limits, Google Upgrades Long-Context Agents●2026.08.24Open Source AI Alliance Reshapes Global LLM Power Dynamics→2026.08.17AI Arms Race Reshapes Global Power: How LLM Breakthroughs Redefine National Competitiveness→2026.08.10AI Industry Shifts From Open Source to Domestic LLMs and Traffic Control→2026.08.03AI Commercialization Wave: Big Tech vs Startups Vie for Market→
The Week in One Line · One line
“Greenberg Traurig Deploys Agentic AI for Legal Work, Bill Gates Urges AI Limits, Google Upgrades Long-Context Agents”
Two issues a day. Ten minutes to turn AI noise into judgment — mornings for the world, evenings for China.
人民公园说AIJustSayAI Weekly Report
JustSayAIWeeklyReport
Two issues a day — AI noise into judgment.
Next Issue · next
2026.09.07 · every Monday →