Results
1,543 articlesSort
0:43ENA Cultured-Neuron Interface Controls DoomMatthew Berman · Forward FutureMatthew Berman describes a Cortical Labs demonstration that connects activity in cultured human neurons to actions in a video game.
2:06:23ENFive AI Employees Live on Five Mac MinisHost Nathan LabenzAfter selling her YC packaging company, Jesse Genet opened a terminal for the first time and built an OpenClaw staff that now plans lessons, logs learning, and keeps family evenings from sliding into slop.
1:49:53ENDon't Fight Backprop: Goodfire's Bet on Designing What Models LearnHost Nathan LabenzDan Balsam and Tom McGrath return to The Cognitive Revolution with a $150 million Series B, a new research pillar called intentional design, and a rule for anyone who wants to steer training: stop trying to out-muscle gradient descent.
1:18ENNVIDIA’s Alpamayo and the Promise of Simulated Driving DataMatthew Berman · Forward FutureMatthew Berman discusses NVIDIA’s autonomous-driving tools and a Mercedes demonstration, while predicting wider adoption.
10:43ENMyFitnessPal's Cal AI Acquisition and the Case for Personal AI AgentsMatthew Berman · Forward FutureMatthew Berman builds a food-photo tracking prototype and argues that agents could threaten standalone apps. The demo also reveals the gap between a quick prototype and a proven product.
1:28ENMatthew Berman on Cal AI, MyFitnessPal and the Cost of Replicating SoftwareMatthew Berman · Forward FutureThe acquisition of a teen-built calorie tracker prompts a debate about what AI coding can replicate—and what remains difficult to build.
0:24ENA Surprise in AI Token Usage and CostsMatthew Berman · Forward FutureMatthew Berman reacts to a rising usage bill and notes why monitoring token consumption matters.
0:18ENA Quick Test of a Generated Python Tetris GameMatthew Berman · Forward FutureMatthew Berman asks for a Python Tetris game and briefly checks whether pieces appear and a line can be completed.
2:17:36ENCurrent Safety Techniques Will Not Give You Many Nines—and They May Fail TogetherHost Nathan LabenzUK AISI chief scientist Geoffrey Irving on jailbreaks that never fail, correlated alignment failures, and why a government lab is funding theory.
0:48ENCheaper Code and Software Jobs: Matthew Berman on the Rebound EffectMatthew Berman · Forward FutureBerman argues that lower software production costs can create new demand. Indeed's job-posting data adds context but does not establish that AI caused a hiring rebound.
0:28ENBaby Keem Asks About an OpenClaw IssueMatthew Berman · Forward FutureMatthew Berman reacts to a public question about internal reasoning appearing in OpenClaw output.
0:42ENNano Banana 2 and World-Knowledge Image GenerationMatthew Berman · Forward FutureMatthew Berman highlights Google's Nano Banana 2, including text rendering, infographic styles and high-resolution output.
13:28ENAnthropic’s Distillation Allegations and the Debate Over AI Training DataMatthew Berman · Forward FutureMatthew Berman examines Anthropic’s allegations against three Chinese AI labs, the criticism of its own training practices, and the geopolitical claims surrounding the dispute.
0:44ENMatthew 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.
1:16ENAnthropic’s Distillation Allegations Draw CriticismMatthew Berman · Forward FutureMatthew Berman examines reactions to Anthropic’s claims about model distillation and debates over training data.
2:03:56EN260 Doctors and 49,000 Criteria: How OpenAI Built Attending-Level Health AIHost 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.
40:00ENHow Matthew Berman Built an OpenClaw Workflow for His TeamMatthew Berman · Forward FutureA tour of sponsorship triage, CRM, meeting notes, security, scheduling, memory and backup after a month of heavy use.
54:26ENTen Billion Active Parameters, and the Unsexy Work of Making Them Agent-ReadyHost Nathan LabenzMiniMax researcher Olive Song explains how M2 was trained with expert developers as reward models, interleaved thinking, and a perturbation pipeline that treats every scaffold change as the real generalization test.
0:25ENGemini 3.1 Pro: Four Benchmark Results from Its 2026 LaunchMatthew Berman · Forward FutureMatthew Berman reviews reported academic, abstract-reasoning and coding scores; the chart distinguishes ARC-AGI-2 from GPQA Diamond.
6:22ENGemini 3.1 Pro: Benchmarks, Animated SVGs, and CAD DemosMatthew Berman · Forward FutureMatthew Berman reviews Google’s February 2026 model release, comparing reported benchmark results and examining examples of generated animations, an urban-planning app, and CAD code.
0:27ENGemini 3.1 Pro Demonstrations: Flocking and City PlanningMatthew Berman · Forward FutureMatthew Berman looks at two visual simulation examples presented with Gemini 3.1 Pro.
0:37ENGemini 3.1 Pro: A Look at the Reported Benchmark ScoresMatthew Berman · Forward FutureMatthew Berman reacts to a Gemini 3.1 Pro comparison and highlights results on Humanity’s Last Exam.
10:17ENAnthropic Subscription Credentials and OpenClaw: A 2026 Policy DisputeMatthew Berman · Forward FutureMatthew Berman explains why restrictions on Claude subscription OAuth credentials prompted him to change his OpenClaw model setup.
0:18ENGemini 3.1 Pro’s Pelican and Frog SVG ComparisonsMatthew Berman · Forward FutureMatthew Berman reacts to two side-by-side animated SVG examples.
Load moreShowing 24 of 1,543









