Browse Comments — Raw (as collected)

Revise your curriculum, every university must inlcude AI courses like its a mandatory mathematics. It should be done through government ministries to re,evaluate all fields of studies. Otherwise there will be less humans ready for these new 1.7 M jobs.

Great point, Paul. The shift from model capability to workflow integration is a natural evolution. Minimizing tool switching allows teams to focus entirely on execution rather than managing disjointed subscriptions.

I think this is spot on, Paul Storm! For me, saying “RIP ChatGPT” isn’t about the tool dying, it’s about finally moving beyond one-off prompting to real systems and better orchestration. As I say in my keynotes, the breakthroghs come when we stop using AI like a fancy search box and start building with it

There is something practical about platforms trying to reduce friction instead of only competing on features or model benchmarks. People usually work better when ideas can move more smoothly from thought to execution.

Shift away from standalone tools toward unified environments reflects a broader pattern in software history. Integration usually solves friction on surface level while redistributing complexity underneath. Users often gain speed, yet lose some visibility into how outputs are assembled across models.

Value in modern AI ecosystems is increasingly tied to end-to-end workflow design, Paul. Efficiency gains tend to emerge when multiple capabilities operate within a single environment.

I wonder if the outcome would have been different if the prompt had specified that the patient was an expat living in the U.S. To me, the model’s behavior seems fairly logical. If a symptom description is written in Japanese, Chinese, or Hindi and no location is provided, the most likely assumption is that the patient is located in a region where that language is commonly spoken. Healthcare systems, care pathways, and thresholds for recommending the ER vary significantly across countries. This becomes even more interesting with languages that are spoken across many regions. French may point to France, Belgium, Switzerland, Quebec, or several African countries. Spanish could mean Mexico, Argentina, Spain, Colombia, or many others. Even English spans countries with very different healthcare practices. The real question may not be whether the model is culturally profiling the user, but whether it should be making geographic assumptions at all. In cases where location materially affects the recommendation, asking for location first might be the safer approach.

People rarely leave tools they outgrow, they leave the friction between tabs, prompts, and unfinished work.

Yeah I have to build a compute mesh connecting all of my neighbors and their compute together. If I connected all gaming computers in the world together AND invented an architecture to run efficiently on it that’s about 100x the worlds top 100 datacententers. Then you could simulate a human brain and shit. And gaming like this would be trivial - just need to invent the infrastructure and get people to trust eachother enough to integrate the economy of the system…

The consolidation angle is what makes this interesting, not just the features. Paul Storm

Removing friction is the name of the game in AI. It's all about making things faster and easier.

Every major technology shift follows the same pattern. People first dismiss it, then mock it, then quietly realize the people using it effectively are moving faster than everyone else.

The plug it in bit…I work with lots of clients who understandly aren’t comfortable connecting their email and Drive to an LLM. Not yet anyway. I wish more was written about what this actually means and how to safeguard privacy and client confidentiality (something on my list)

The fatigue from constantly switching platforms is real and often overlooked. Unifying the process could change how fast ideas actually turn into output.

It is interesting how the focus is moving from model quality to execution flow. The winners will likely be the ones who make creation feel seamless end to end.

Most people do not need more tools, they need less friction between steps. Consolidation like this could reshape how everyday AI work actually gets done.

AI value increasingly depends on how seamlessly capabilities connect across the workflow, not simply how many models are available, Paul.

The real personal operational bottleneck isn't an absolute deficit of frontier model processing capacity or baseline algorithmic intelligence, but an individual habit of mistaking a fragmented, multi-tab execution strategy for a viable long-term productivity framework.

this feels like a shift from model comparisons to how quickly people can move from idea to output across tools, Paul

Smart move, Paul. Consolidation is where AI is heading.