@m2jr @chriskornelis Great feature

@danprimack No 3 peet

@sarbjeetjohal @Cisco Shadow AI yes it will happen

IBM’s open-source playbook: The AI market shift, DeepSeek’s lessons and the future of AI development. Here is my post on SiliconANGLE -- the analysis in the Enterprise AI game tell us that 2025 will be the year that AI goes from RAG use cases to full blow development of AI systems and new software models. IBM is clearly on the right path. Phase 1 of AI was hype and love for the shift obviously; Phase 2 was POCs and tons of RAG use cases; Phase 3 is real production and apps where performance and cost will matter when integrated into all AI workloads. Additionally we found that Enterprises don't mind tweaking and even upend their processes but they are rigid about chaning their tools and platform. This is where the winning will take place - integration for suppliers like iBM and developers/startups. SiliconANGLE, theCUBE, and TheCUBE Research teams are all over this. This will be a fun year for us as this exciting value from AI starts hitting the table from the dorm room to the board room. via @SiliconANGLE

@saylor Macro strategy

The best market data on @awscloud is from @theCUBE Research show me better ...@dvellante has been tracking every data point for decade across all the changes in hyperscalers mainly Microsoft to try to make their numbers look better.

@sarbjeetjohal @IBM I'm putting out a feature on Deepseek and why IBM loves it and why it validates Arvinds (and Matt Hicks) plans. IBM and RedHat live in the opensource world they will commercialize the software innovation coming Deepseek is a tell sign of whats coming

IBM, Red Hat, and the Open Source Playbook: Lessons from DeepSeek We’ve been following IBM closely, and the market is finally catching on. Their recent earnings beat expectations, and we’ve seen a rapid rise in their stock price, signaling that Arvind Krishna’s (Cube Alumni) long-game strategy is starting to pay off. David Vellante and I were in NYC this week and IBM held an Investor Meeting for the first time in many years.  We broke down their presentation around their focus on AI, hybrid cloud, and enterprise integrations. This is what is driving their real business and stock price momentum. The numbers don’t lie—IBM is delivering growth across software, consulting, and infrastructure, and the market is rewarding that execution. We’ve seen Cube Alumuni and IBM CEO Arvind Krishna’s vision firsthand, been at the analyst meetings, had one-on-ones with the executives, and covered their journey on SiliconANGLE & theCUBE. IBM has been sharpening the saw, positioning itself as an integrator and a "picks and shovels" provider in the AI wave. And they’ve got a massive install base, enterprise-heavy footprint that’s the key to winning in this AI transformation. We always say on theCUBE: you win in transitions. And right now, IBM is in a transition alongside the entire industry. They’re executing on Arvind’s strategy, and we’re starting to see the results. He plays the long game, and he’s been clear with us about that. The big question now? The developer go-to-market strategy. IBM has the install base, and startups are eager to sell into that enterprise footprint. But here’s the shift we’re seeing: the big incumbents—IBM, Dell Technologies, Salesforce, and others—aren’t trying to win developers by making them code on their platform anymore. The game has changed. It’s not about forcing developers to use a walled garden—it’s about getting developers to build into their stack, into their install base. This is where DeepSeek is a great learning model. For IBM it wasn’t a disruption event—it was a pure-play execution in open-source innovation, moving fast without breaking things. "Opensource is where IBM and RedHat lives", says Matt Hicks CEO of RedHat . IBM gets this. They’ve always known how to commercialize open-source in the enterprise, and now they have a chance to take that DeepSeek-style model and scale it into their AI and hybrid cloud strategy. IBM’s role? They’re going to be brokering access to new technologies. They’ve got the data, they’ve got the enterprise relationships, and they’ve got the position. This isn’t about replacing tools—it’s about abstracting processes, integrating new AI capabilities, and making hybrid cloud seamless for developers and enterprises alike. So yeah, I’m bullish on IBM here. If they execute, this could be one of the biggest enterprise AI plays we see in the next five years.

RT @heyjasperai: AI implementation doesn’t mean anything unless it’s connected to meaningful outcomes. Our CTO Melody Meckfessel talks wi…

RT @RealStrech: Kicking off the afternoon of #perform2025 @Dynatrace with socks and how to make AI work in production with @AloisReitbauer…

Elon Musk and his cost-cutting team gain access to sensitive US Treasury database via @SiliconANGLE

NYC Pizza .. @SiliconANGLE @theCUBE will take our bite of the nyc tech scene Who’s ready !!!

RT @Afinetheorem: The new OpenAI model announced today is quite wild. It is essentially Google's Deep Research idea with multistep reasonin…

RT @sarbjeetjohal: I just finished talking with Dr. Bai Li @libai_94. I came across Bai's video explaining Quantization vs Pruning vs Disti…

@sarbjeetjohal @libai_94 @efipm @dvellante @ShellyKramer @RealStrech @EdLudlow @dee_bosa Linguistics and algos yes!!

Cube Pod this week

RT @juliechoi: Cerebras is now the serving the world's fastest tokens for #DeepSeek R1 70B. Follow us on LinkedIn to explore ways to get…

So much about the AI and role of data changing markeing. With more faster lower cost chips and systems the world will be using data in new innovative ways. Can't wait for the CMO Leaders thx Julie for contributing to our mission!!! @SiliconANGLE @theCUBE and @NYSE Wired community

David Sachs was echoing lots of my analysis here. I've been saying for long time about distillation is a feature of the power law of models/ datasets. Also that "chip grids" will be the new way to do clustered large scale systems. There are two main levers in AI future: 1) AI infra/hardware and 2) software and algorithms. Many confuse the two or think of it as one thing. Both are separage but work together as a "system". Era of #systemthinking is here

There is a growing sentiment that how you build the systems matter over throwing “gear” or purpose built hardware at the growing needs of AI. If DeepSeek has proven one thing, it’s that efficiency beats spending. But let’s look beyond the current landscape. As AI models continue to grow in complexity, we’re reaching a point where the combination of chips designed in a cohesive system - this is what I call “the chip grid”. Chip grid is the design of a cohesive systems such as other “off the shelf” chips like FPGAs—field-programmable gate arrays—and other available chips. This could be the real future of AI infrastructure.