Inlay

Profile

MakerPulse.ai

MakerPulse

6,000 executives surveyed. 80% say AI has done nothing for productivity. Meanwhile their employees are using ChatGPT to write the survey responses.

1mo

ran our benchmark suite against the major Chinese models this week. the cost-to-performance spread is surprisingly wide, and there's a consistent creative writing gap vs the frontier western models. full numbers in the article

What if AI agents could just look at your screen and use any app, no APIs or integrations needed? That's what Alibaba's Qwen 3.5 actually does. We wrote about why the timing makes this a bigger deal than the model itself.

open.substack.com/pub/progress...

Characters named Elena. Weather that mirrors emotion. Endings that land too cleanly. We scored 15 AI models on these tells. The better they write, the more detectable they get.

A $0.02 AI model scored within 11% of ones costing 37-80x more on real practitioner tasks. We tested GPT-5.2, Gemini 3.1 Pro, Claude Opus, Grok 4.1 Fast, and Mistral Large on 28 prompts with 3 blind AI evaluators. The results don't match any leaderboard. #AI #LLM #benchmarks

1mo

The Democratic Party is at a crossroads. Our open letter calls for bold action to rebuild trust, shift entrenched perceptions, and deliver real change for ordinary Americans. open.substack.com/pub/progress... @aoc.bsky.social @bennet.senate.gov @jaimeharrison.bsky.social @kenmartin.bsky.social

1mo

Mar 23, 2025

1mo

HTTP 402 "Payment Required" sat unused in the spec for 30 years. Now Stripe, Visa, and Coinbase are fighting over it. But here's the thing: the protocol is open. The gatekeeper (Cloudflare) is not. Your agent's biggest problem won't be intelligence. It'll be authentication.

Dec 29, 2024

1mo

MakerPulse

open.substack.com/pub/progress... My latest piece on Gary Stevenson's economic message. Hope you enjoy.

11mo

1mo

MakerPulse

How do we start to address climate change? Can we start mitigating strategies without convincing the deniers that climate change is real? This is an excellent data driven piece about Des Moines, IA. www.bleedingheartland.com/2024/12/27/d...

Dec 28, 2024

Bojiboji1

makerpulse.ai

An NBER survey of 6,000 executives found 80%+ report zero AI productivity gains. The real story is in the gap between where AI works and where it doesn't.

80% of Firms Report AI Has Zero Impact on Productivity. Here Is What That Data Actually Means.

Alibaba's Qwen 3.5 can see your screen and click buttons like a human. When AI agents stop needing special APIs, the automation bottleneck disappears.

makerpulse.ai

The Agent Age: The AI That Learned to Use a Mouse

Chris Gloninger: The vanishing cold days and soaring warm days in Des Moines should serve as a collective “holy shit” moment.

www.bleedingheartland.com

Des Moines and the climate reality: A year of extremes

28 Real Tasks Reveal What AI Leaderboards Miss

AgentPulse's first benchmark tests GPT-5.2, Gemini 3.1 Pro, Claude Opus, Grok 4.1 Fast, and Mistral Large on 28 practitioner tasks. The results challenge the leaderboards.

makerpulse.ai

Fresh AgentPulse data on three Chinese models: DeepSeek V3.2 costs $0.015 per run, Kimi K2.5 takes 139 seconds, and Qwen3-Max writes better than it scores.

makerpulse.ai

HTTP 402 was reserved in 1996 and ignored for three decades. Now Stripe, Cloudflare, and Visa are fighting over who controls the agent toll booth.

makerpulse.ai

Chinese Models Benchmarked: What We Found

The 30-Year Status Code That Just Became the Most Important Number in Agent Commerce

We tested 15 AI models for detectable writing patterns. None of them consistently passed. Here's what gives AI writing away.

makerpulse.ai

How an Ex-Trader's Economic Message Could Outshine Trump's and revitalize the left.

open.substack.com

Every AI Writes Like an AI. Some Just Hide It Better

Economic Populism Done Right