AI Arms Race Volume: 3
Author: Brod Justice, Ryan McClure
Published Date: August 8, 2025
Weekly AI Signals — Curated by Builders, for Builders
Every week, we round up the most relevant developments in AI, tech, and the business strategies shaping them. Born from internal conversations among senior engineers, product leads, and AI researchers, this digest simplifies the noise into clear signals. Authored by Brod Justice and Ryan McClure, these summaries are built for real practitioners—people shipping code, scaling platforms, and thinking past the hype.
Volume 3
1. Trump Partners with “Content-Stealing” Perplexity AI
Trump has signed a deal with Perplexity AI, the content stealing chatbot. A cynical person might link this with Trump’s announcement last week suggesting that copyright laws should be ignored: PR here: amazonaws.com Interesting info from TechCrunch here: techcrunch.com
2. EU’s Landmark AI Act Takes Effect—But Falls Short for Creators
The EU’s AI Act came into effect this week. It’s big, and we certainly haven’t digested all of it. While we support the intent, we currently see its practical protection for content creators to be very limited. It does not even get close to replacing the technical solutions we are driving: https://artificialintelligenceact.eu/ai-act-explorer/
3. Krugman: AI Boom Masks Trump’s Economic Missteps
Is the AI boom in the USA disguising Trump’s economic policy errors? Some seem to think so, including economist Paul Krugman who claims it accounted for about half of the USA’s growth in 2025. paulkrugman.substack.com
4. OpenAI’s GPT-OSS: A Scorched-Earth Move Against China’s Open Source Models
OpenAI released a very good OpenSource model, GPT OSS. This seems like a scorched-earth policy: it’s cheaper, faster and sized to hit the OpenSource Chinese models. But it also hits Meta, Grok, Mistral, etc. One drawback, it’s primarily an English-only model, our own German tests show it to be somewhat clumsy: openai.com
5. GPT-5 Impresses—But Still Fails the “Random Number” Test
OpenAI also released GPT5, to great fanfare. While it seems to have many improvements (its coding abilities my take customers away from Anthropic Claude/Opus), it’s not a great leap forward and it still fails our simple “random number test”: linkedin.com
6. OpenAI Coders Still Prefer Claude 3.5 Over Their Own Models
OpenAI developers are apparently using the old Anthropic models rather than their own. This actually makes sense. We know from our experience that Claude 3.5 (which is over a year old at publication) is better at coding than anything from OpenAI (though see also point 5 above). More evidence that in some cases AI “intelligence” has already peaked: wired.com
7. Elon Musk Says OpenAI Will “Eat Microsoft Alive”
Elon Musk warned that OpenAI could soon “eat Microsoft alive,” referring to the release of GPT‑5 and its disruptive potential—even to a major partner like Microsoft.
Source: news.com.au news.com.au
8. OpenAI Reinstates GPT-4o After User Backlash
Following negative user feedback, OpenAI reinstated GPT‑4o for ChatGPT Plus subscribers. The model’s more personable tone had been missed compared to the newer GPT‑5.
Source: Windows Central windowscentral.com
9. Reddit Users Slam GPT-5 as a Downgrade
Nearly 5,000 users flocked to Reddit, arguing that GPT‑5 “feels like a downgrade” due to shorter, less informative responses and more limited interaction compared to prior versions.
Source: Tom’s Guide tomsguide.com
10. AGI Still Out of Reach, Despite GPT-5 Hype
An insightful piece explores the accelerating race toward AGI. While GPT‑5 moves us forward, it still lacks core autonomous capabilities—keeping true AGI as an ambitious, elusive goal.
Source: The Guardian theguardian.com
11. GPT-5 Launch Marred by “Chart Crime” in Reddit AMA
During an August 8 Reddit AMA, OpenAI CEO Sam Altman admitted the GPT‑5 launch had a “bumpy” start—and referenced a viral “chart crime” incident that sparked confusion. Some users reverted to GPT‑4o as a result.
Source: Times of India timesofindia.indiatimes.com
12. AI Bot Traffic Trends: Publishers Push Back
AI-powered crawlers are scraping entire articles, cutting into publishers’ revenue. Arc XP’s DataDome integration lets them detect, block, or even monetise bot traffic.
13. Economist: AI Is Killing the Web
AI answer engines are bypassing source sites, causing a sharp drop in referral traffic and ad revenue. The Economist warns creators must adapt fast with paywalls, licensing, or new formats.
New to AI? Take Our Free Course.
ChatBar AI Learning is the fast, practical way for business owners and teams to get up to speed on real-world AI.
Want More Like This? Just Ask.
Tap below and ask ChatBar to recommend more high-signal AI content and use cases—no fluff, no noise.
Ask ChatBar to show similar content
Ready to Try ChatBar AI on Your Site?
Apply now to join our early access program and see how ChatBar can supercharge your site’s content, discovery, and engagement.