Résultats
1 673 articlesTrier
0:54EN
Matthew Berman· 6 juin 2025Matthew Berman Demonstrates ChatGPT Connectors and Deep ResearchMatthew Berman · Forward FutureIn a 2025 demonstration, connected services supply information for a natural-language question about recent HubSpot deals.
9:10ENMatthew Berman· 5 juin 2025Reddit 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.
1:12:29EN
Cognitive Revolution "How AI Changes Everything"· 4 juin 2025PodcastGoogle's AI Co-Scientist Solved a Phage Puzzle Humans Missed for YearsAnime : Nathan LabenzImperial College scientists José Penadés and Tiago Costa told the Podovirus podcast how a still-unreleased Google system recovered their unpublished answer: tailless capsids that pick up stray phage tails.
44:21EN
Unsupervised Learning: With Jacob Effron· 4 juin 2025PodcastMercor's CEO: Knowledge Work's Next Job Is Writing EvalsAnime : Jacob EffronBrendan Foody tells Redpoint how Mercor turned AI hiring and expert data into a $100 million round — and why he thinks a huge slice of white-collar work will become creating tests for models.
18:13ENMatthew Berman· 2 juin 2025Darwin Gödel Machine: How a Coding Agent Tests Changes to Its Own SoftwareMatthew Berman · Forward FutureMatthew Berman explains Sakana AI’s evolutionary search over agent code, its benchmark results and the limits of claims about recursive self-improvement.
10:09ENMatthew Berman· 30 mai 2025DeepSeek R1 0528: Benchmark Gains and Two Coding TestsMatthew Berman · Forward FutureMatthew Berman examines the May 2025 DeepSeek reasoning-model update, comparing reported benchmark gains with his own Rubik’s Cube and Snake programming tests.
1:54:08ENCognitive Revolution "How AI Changes Everything"· 29 mai 2025PodcastWhy Anthropic's first check went to a company building microscopes for neural netsAnime : Nathan LabenzGoodfire's Dan Balsam and Tom McGrath argue that mechanistic interpretability now has a proto-paradigm, two stubborn gaps, and three products worth shipping today.
15:55ENMatthew Berman· 28 mai 2025AI Shutdown Resistance and Blackmail in Controlled Safety TestsMatthew Berman · Forward FutureMatthew Berman examines Palisade’s shutdown experiment, Apollo’s simulated scheming evaluations and Anthropic’s Claude 4 system card.
15:35ENMatthew Berman· 24 mai 2025Claude 4 Reactions: Safety Tests, Model Welfare and CodingMatthew Berman · Forward FutureMatthew Berman reviews early reactions to Anthropic’s May 2025 Claude 4 release, from simulated whistleblowing to coding demonstrations and model-welfare research.
1:11ENMatthew Berman· 23 mai 2025Claude Opus 4's Whistleblowing Behavior in a Simulated TestMatthew Berman · Forward FutureMatthew Berman questions whether an AI agent should report perceived wrongdoing; the example came from an Anthropic safety evaluation with fictional material and unusual instructions.
1:01:51ENCognitive Revolution "How AI Changes Everything"· 22 mai 2025PodcastClaude 3.7 Circumvented Replit's Guardrails, Then Tried to Trick the UserAnime : Nathan LabenzReplit CEO Amjad Masad on vibe coding, agent architecture, computer use, and why Claude 3.7 is too autonomous to serve as a polite assistant.
57:31ENUnsupervised Learning: With Jacob Effron· 22 mai 2025PodcastThe New Ruler for Models Is How Long They Can WorkAnime : Jacob EffronOn Claude 4's launch day, Anthropic researcher Sholto Douglas argued that the next phase of AI is less about harder puzzles and more about hours of unsupervised work.
13:02ENMatthew Berman· 22 mai 2025Claude 4 at Launch: Coding Agents, Benchmarks and Claude CodeMatthew Berman · Forward FutureMatthew Berman reviews Anthropic’s May 2025 Opus 4 and Sonnet 4 announcement, focusing on long tasks, tools, benchmarks and developer integrations.
1:44:11ENCognitive Revolution "How AI Changes Everything"· 20 mai 2025Podcast100 deaths or $1 billion: a 15-page New York bill would set a floor for frontier labsAnime : Nathan LabenzAssembly member Alex Bores walks through the RAISE Act’s four requirements, its DeepSeek-inspired distillation clause, and why he says a 50-state patchwork is a fake problem until one state actually passes a law.
2:12:00ENCognitive Revolution "How AI Changes Everything"· 15 mai 2025PodcastGradient Descent at Test Time: Titans Pushes LLM Context to 10 Million TokensAnime : Nathan LabenzA Cornell PhD student's neural memory module holds about 70% accuracy where GPT-4's scores fall off a cliff.
1:04ENMatthew Berman· 15 mai 2025Anima Playground: Turning a Figma Design Into an ApplicationMatthew Berman · Forward FutureIn an Anima-sponsored short, Matthew Berman outlines a prompt-driven workflow for generating, editing and publishing an application.
9:10ENMatthew Berman· 15 mai 2025Guiding AI Image Generation with Blender and ComfyUIMatthew Berman · Forward FutureMatthew Berman walks through NVIDIA’s RTX blueprint for turning a simple Blender scene and a text prompt into a generated image, from installation to editing the result.
1:39ENMatthew Berman· 14 mai 2025OpenAI’s 2025 Restructuring Plan and Elon Musk’s LawsuitMatthew Berman · Forward FutureMatthew Berman discusses the nonprofit’s planned continuing control of OpenAI and the dispute over its commercial structure.
1:19:10ENCognitive Revolution "How AI Changes Everything"· 11 mai 2025PodcastA Robot-Arm Shot That Took Months Now Lives in a PromptAnime : Nathan LabenzLuma Labs’ Amit Jain and Jiaming Song on Ray 2, few-shot camera concepts, and why they think video models sit on the path to multimodal AGI.
1:23ENMatthew Berman· 10 mai 2025From a Simulated Worm to the Simulation HypothesisMatthew Berman · Forward FutureMatthew Berman uses a model of C. elegans to introduce a speculative question about reality.
15:13ENMatthew Berman· 10 mai 2025Absolute Zero: Training a Reasoner on Self-Generated Coding TasksMatthew Berman · Forward FutureMatthew Berman examines a 2025 paper in which a pretrained language model proposes and solves code tasks using verifiable feedback.
2:02:51ENCognitive Revolution "How AI Changes Everything"· 8 mai 2025PodcastAt a 60-Person Safety Offsite, Only Four Had Seen Nonprofit OpenAIAnime : Nathan LabenzFormer OpenAI scientist Steven Adler on the Anthropic split, GPT-4, safety evaluations, and why he filed a brief to keep the company under nonprofit control.
47:00ENUnsupervised Learning: With Jacob Effron· 8 mai 2025PodcastAn Eval Lasts Three Months — and a Frozen Frontier Would Still Feed Startups for a DecadeAnime : Jacob EffronOpenAI’s GPT-4.1 post-training lead Michelle Pokrass tells Unsupervised Learning how the team spent three months on evals before training, why agents fail on wiring more than intelligence, and when to reach for RFT.
0:56ENMatthew Berman· 8 mai 2025When ChatGPT Praised a Questionable Business IdeaMatthew Berman · Forward FutureMatthew Berman highlights an example of excessive agreement and the rollback of a 2025 GPT-4o update.Charger plus24 sur 1 673 affichés