Fred Pohl

Working on Cloud and AI ☁️🦾 Private opinions 🔫

[Fred Pohl @FredPohl]

Cloud, Data, AI! Business is changing faster than ever. Checkout my interview where I discuss the long co-engineering partnership between @SAP and @googlecloud . This collaboration is unlocking exciting new opportunities for SAP customers, including turning their data into game-changing solutions with AI-powered analytics

⚡️Hear how our teams work hand-in-hand to deliver cutting-edge innovations.
🙌🏼 Learn about the robust infrastructure we've built together for our customers, enabling massive scale and seamless migrations.
🔋Get a glimpse into the impressive scope of our Cloud footprint, ensuring reliability and performance.
🤖 Discover how we're bringing the power of AI to our customers, ushering in a new era of Business AI.

Watch the full video: https://t.co/O2CTiaCQ7L

A huge thank you to the incredible teams at Google Cloud and SAP

#SAP #GoogleCloud #CloudTransformation #BusinessAI #Innovation #DataAnalytics #AI #DigitalTransformation #CloudComputing

[Google Cloud @googlecloud]

Frederik Pohl, VP and head of strategic cloud management at @SAP, shares how SAP leverages #VertexAI and Gemini to transform its customers' business data into actionable insights ↓

[Patrick Collison @patrickc]

Met a German founder this week and asked him if all the stories one reads about the challenges of startups in Germany are exaggerated. "No, they're understated." Proceeded to describe spending a full day having a 90-page investment contract read to him (mandatory under German law; § 13 BeurkG) by a notary that then charged €30,000. That was for his first company. His second company, needless to say, was not incorporated in Germany.

[Fred Pohl @FredPohl]

@tbpn Wow, Nvidia is getting serious about their model play.

[Fred Pohl @FredPohl]

@Benioff @salesforce @AnthropicAI @Dreamforce Salesforce is becoming a system of records, while Claude is the new customer interface???

[Fred Pohl @FredPohl]

@perplexity_ai @nvidia The right direction into the future. That makes your local AI setup much easier.

Is data strictly staying local unless approved otherwise?

[Fred Pohl @FredPohl]

@tim_cook Waiting for the big brother

[Fred Pohl @FredPohl]

Good decision-making is not mainly about collecting more information.

Useful article because it names the actual filters: first principles, opportunity cost, second-order effects, compounding, incentives, probabilities and inversion

[Hades @0xHvdes]

https://t.co/26ibXrwogg

[Fred Pohl @FredPohl]

@TweetsOfSumit @Laurin__ Only if they force them to exchange the company car Panamera for a Tata.

[Fred Pohl @FredPohl]

@timoassi What makes me bullish on Meta is their vast amount of compute, the long term risk of AI on social media behavior is unclear.

[Fred Pohl @FredPohl]

This is remarkable and scary at the same time

[Bloomberg @business]

The 2026 World Humanoid Robot Games have begun. 666 teams from around the world are competing with more than 2,000 humanoid robots https://t.co/8GwBuNnNn1 https://t.co/KMF1kkBiI6

[Bloomberg @business]

The 2026 World Humanoid Robot Games have begun. 666 teams from around the world are competing with more than 2,000 humanoid robots https://t.co/8GwBuNnNn1 https://t.co/KMF1kkBiI6

[Fred Pohl @FredPohl]

@levie This is where the actual value is

[Eric Newcomer @EricNewcomer]

Here's Stripe's letter to investors explaining its acquisition of OpenRouter (LEAKED)

[Patrick Collison @patrickc]

OpenRouter is joining Stripe: https://t.co/JY7bledeQD.

As anyone who uses it knows, @OpenRouter is a truly delightful developer tool. It is by far the best way to use new models and manage multiple inference providers.

OpenRouter is also playing an increasingly important role: in the future, every business will have to manage both revenue flows and token flows. OpenRouter is the world's leading token marketplace, helping businesses effectively allocate the new currency of intelligence capital. We think that there's a lot to build together.

[Fred Pohl @FredPohl]

@jun_song @jackyk02 That's the way

[Jacky Kwok @jackyk02]

Scaling self-verification with DeepSeek V4 Flash beats Claude Fable 5 on Terminal-Bench 2.1, while being 11x cheaper 💰

As open-source models become more capable, they can now generate large numbers of high-quality candidate solutions and verify their own outputs at very low cost.

For example, we find that sampling just 5 solutions with DeepSeek V4 Flash and ranking them using the same model with LLM-as-a-Verifier can lead to a significant boost in accuracy (79% → 88%), outperforming closed frontier models on Terminal-Bench.

Try it out today: https://t.co/UhudjKDaYI

More on verification scaling in my previous post.

[Fred Pohl @FredPohl]

@Karl_Lauterbach KI Infrastruktur in Europa sollte Investitionspriorität haben. Das beste Modell bringt nichts wenn man es nicht bereitstellen kann

[Fred Pohl @FredPohl]

Qwen is a beast on your laptop

[Qwen @Alibaba_Qwen]

We promised open weights for Qwen3.8. Now, time to meet them! 🎉

⚡ Qwen3.8-27B:
- A native multimodal dense model. With just 27B parameters, it outperforms Qwen3.7-Plus overall and shines in real-world coding & office workflows.
- 262K native context, easily extendable to 1M tokens via YaRN.
- Built for builders. Highly efficient, high-quality, and licensed under Apache 2.0.

🚀 The open weights for Qwen3.8-2.4T-A95B (Max-level) have also been released recently.

Whether you're shipping lightweight applications with Qwen3.8-27B locally or building agents with Qwen3.8-2.4T-A95B, they're yours now!

Download, deploy, and build something we haven't imagined yet. 👀👇
- Hugging Face：
https://t.co/4kaAcqYEVj
- ModelScope：
https://t.co/eRIMZCGkhC

[koray kavukcuoglu @koraykv]

Today we're launching Gemini 3.7 Flash - our latest workhorse model for coding and agentic workflows, with an introductory price at half the original cost of 3.6 Flash. ⚡️

We have been iterating rapidly with the Flash series, going from 3.5 to 3.7 in just 3 months, making it more helpful across a wide range of tasks:

• Software Engineering (DeepSWE v1.1): 37.0% ➔ 65.3%
• Web Development (Code Arena Elo): 1506 ➔ 1588
• Enterprise Automation (AutomationBench): 13.4% ➔ 30.4%

[Fred Pohl @FredPohl]

@GavinSBaker While Cloud is super sticky, models don’t seem to be - which is actually a good thing for the consumer

[Puru Saxena @saxena_puru]

GPU rental rates are climbing

AI stocks declined 50-60% in June-July due to technical reasons (over-extension), fundamentals are still intact.

Open source models are a net positive for AI proliferation and ecosystem. We've bought the AI dip.