最近更新されたチャンネル

選ぶとそのチャンネルだけを表示すべてのチャンネル →

絞り込み結果

39 件
並び順
スライド 3 枚13:21EN
Adrian Twarog· 2026年10月21日Adrian Twarog: Giving Enterprise Vibe Coding a Governed Path to ProductionAdrian TwarogA Superblocks walkthrough shows how shared integrations, isolated databases, policy agents and audit logs can bring business-built AI apps under IT oversight.Vibe Codingエンタープライズ AIAI security+2
スライド 3 枚17:41EN
AI Engineer· 2026年10月20日Al Chen: Evidence, Recovery and Data Controls for AI in BanksAl Chen · CiscoA practical discussion of vulnerabilities, disaster recovery and sensitive data in bank AI deployments.エンタープライズ AIセキュリティと安全性Privacy & data governance+2
スライド 2 枚0:36EN
Matthew Berman· 2026年10月13日OpenAI Employee Charged in Recording-Device InvestigationMatthew Berman · Forward FutureMatthew Berman discusses the reported discovery of a recording device and OpenAI’s account of its investigation.セキュリティと安全性OpenAIPrivacy & data governance
スライド 3 枚20:47EN
Matthew Berman· 2026年9月1日Why OpenAI Plans to End Its Direct Model Agreement With CursorMatthew Berman · Forward FutureAfter SpaceX acquired Cursor, OpenAI proposed winding down its model-supply contract. Matthew Berman examines the dispute, competing providers and the risks of relying on another company’s models.コーディングエージェントAI securityAI Strategy+1
スライド 5 枚20:46EN
Matthew Berman· 2026年8月26日Open-Weight AI Models: Token Share, Spending and Strategic TradeoffsMatthew Berman · Forward FutureMatthew Berman examines a Vercel AI Gateway snapshot and argues that model usage and spending can tell very different stories.Open-weight modelsAI cost & performanceAI Strategy+2
スライド 7 枚31:58EN
Theo - t3․gg· 2026年8月14日Claude’s AI Watermarks and the Limits of Detecting Generated ContentTheo Browne · t3.ggTheo Browne examines Anthropic’s marking plans, the challenge of text watermarks, and what signed provenance can and cannot establish.AI SafetyPrivacy & data governanceAI Regulation+2
2:33:35EN
Cognitive Revolution "How AI Changes Everything"· 2026年6月1日ポッドキャストA security researcher audits the AI stack that booked a week of guestsホスト Nathan LabenzNathan Labenz walks Daniel Miessler through a 1GB second brain, two named AI employees, and the blast-radius rules that should govern them.エージェントPersonal agentsエージェント工学+3
スライド 10 枚1:42:41EN
Matthew Berman· 2026年5月1日Testing AI Reasoning, App Store Lookalikes and the Debate Over AnthropicMatthew Berman · Forward FutureIn a May 2026 livestream, Matthew Berman tests a car-wash prompt, discusses AI subscription limits and considers reported employee monitoring for model training.AI agentsBenchmarksPrompting & in-context learning+2
スライド1:11EN
Matthew Berman· 2026年4月24日Anthropic’s Coding Flywheel and the Compute ConstraintMatthew Berman · Forward FutureMatthew Berman sketches a proposed feedback loop linking coding models, enterprise revenue, data and compute, then argues that capacity could slow it.AnthropicAI cost & performanceAI Strategy+1
1:43:39EN
Cognitive Revolution "How AI Changes Everything"· 2026年3月11日ポッドキャストFive Mutations Could Make a 60% Fatal Bird Flu Spread Between Peopleホスト Nathan LabenzJohns Hopkins assistant professor Jassi Pannu argues for locking down only the sliver of functional data that teaches models how to make pandemic viruses — and keeping the rest of biology open.セキュリティと安全性安全性とガバナンスAI security+3
0:44EN
Matthew Berman· 2026年2月25日Matthew Berman on a DJI Robot Vacuum Security FlawMatthew Berman · Forward FutureA custom controller for a DJI Romo exposed a broader authorization problem involving thousands of devices and sensitive household data.CybersecurityPrivacy & data governanceRobotics+1
2:03:56EN
Cognitive Revolution "How AI Changes Everything"· 2026年2月25日ポッドキャスト260 Doctors and 49,000 Criteria: How OpenAI Built Attending-Level Health AIホスト Nathan LabenzKaran Singhal tells Nathan Labenz how ChatGPT Health was evaluated with physicians, walled off from model training, and aimed at raising both the floor and the ceiling of human health.企業と戦略AI in healthcare評価(Evals)+3
スライド1:15EN
Matthew Berman· 2026年2月12日An OpenClaw Personal CRM for Contacts and Meeting PreparationMatthew Berman · Forward FutureMatthew Berman describes a personal workflow that combines email, calendar events and meeting notes to build contact records and prepare for meetings.Personal agentsWorkflow automationAgent memory+1
スライド 5 枚15:10EN
Matthew Berman· 2026年1月24日OpenAI’s Drug-Discovery Value Sharing: What Was Proposed and What Remains UnclearMatthew Berman · Forward FutureMatthew Berman examines Sarah Friar’s remarks on licensing, compares research partnerships and considers a speculative future of concentrated AI resources.AI for scienceAI StrategyPrivacy & data governance+1
スライド 7 枚18:33EN
Matthew Berman· 2026年1月19日OpenAI's Revenue, ChatGPT Go and Ads: A 2026 Strategy AnalysisMatthew Berman · Forward FutureMatthew Berman connects OpenAI’s revenue disclosures, a lower-priced subscription and planned advertising to a wider competition for AI users. The financial estimates and market scenarios are distinguished from company disclosures.Product StrategyPricing StrategyMonetization+1
1:18:40EN
Cognitive Revolution "How AI Changes Everything"· 2026年1月1日ポッドキャストWhen AI Becomes the Oil Well: Luke Drago on the Intelligence Curseホスト Nathan LabenzThe Workshop Labs co-founder tells the FLI Podcast why a world of non-human production could recreate the resource curse — and how private data, open weights, and moonshot careers might still break it.企業と戦略AI Strategy安全性とガバナンス+3
スライド 5 枚14:16EN
Matthew Berman· 2025年10月21日ChatGPT Atlas: OpenAI’s 2025 Browser and Its Agent DemosMatthew Berman · Forward FutureMatthew Berman walks through OpenAI’s launch demonstration of a browser with a ChatGPT sidebar, history search and an agent that can use websites. OpenAI later deprecated Atlas.OpenAIAI agentsComputer use+2
1:32:47EN
Cognitive Revolution "How AI Changes Everything"· 2025年9月13日ポッドキャストA transformer co-author wants a $160 million model no company ownsホスト Nathan LabenzIllia Polosukhin tells Nathan Labenz how NEAR turned a payments headache into a stack for private inference, shared-revenue training, and AI agents that still answer to a court.インフラとデプロイInferenceAI security+3
スライド1:39EN
Matthew Berman· 2025年6月23日Reddit v. Anthropic: Scraping Allegations and the Lawsuit’s ProgressMatthew Berman · Forward FutureMatthew Berman reads from Reddit’s complaint about alleged unauthorized data use; the claims remain disputed as the case proceeds.AI RegulationPrivacy & data governanceData quality & curation+1
スライド 4 枚9:10EN
Matthew Berman· 2025年6月5日Reddit v. Anthropic: The Data-Scraping Allegations and What Happened NextMatthew Berman · Forward FutureMatthew Berman reads Reddit’s 2025 complaint alleging unauthorized access and use of its content. Anthropic disputes the claims, and subsequent procedural rulings have not decided liability.AnthropicPrivacy & data governanceAI Regulation+1
スライド1:03EN
Matthew Berman· 2024年11月29日AnythingLLM’s Community Hub and Local AI Agents on NVIDIA RTXMatthew Berman · Forward FutureIn a sponsored short, Matthew Berman introduces local agent workflows, shared prompts and integrations, with important limits on privacy and performance claims.Local inferenceAI agentsAgent skills+1
57:45EN
Lex Fridman· 2024年11月1日ポッドキャストThe last thing a totalitarian state can take is love — a reread of Orwell’s 1984ホスト Lex FridmanLex Fridman walks back through Oceania’s machinery of Newspeak, Doublethink, and Room 101, and argues that hate-fueled rebellion is not a way out.Political PhilosophyFree SpeechPrivacy & data governance+1
スライド 3 枚8:34EN
Matthew Berman· 2024年7月31日The Promise and Tradeoffs of Running AI on Phones and Other DevicesMatthew Berman · Forward FutureIn a Qualcomm-partnered video, Matthew Berman explores local AI, model routing and demonstrations across phones, PCs, cars and drones.Local inferenceModel routing & gatewaysAI cost & performance+2
スライド 4 枚16:49EN
Matthew Berman· 2024年6月21日How Apple's On-Device and Private Cloud Models WorkMatthew Berman · Forward FutureMatthew Berman examines Apple's 2024 foundation-model overview, from training data and quantization to task adapters and human evaluations.Local inferenceモデルと学習Privacy & data governance+1
もっと見る39 件中 24 件を表示