Jim Scott

Environmental Design, Earth Observation and Geospatial Technology, Exploration Heritage Conservation. CEO-Earth Intelligence.

[Jim Scott @jimscott]

@WillManidis What an incredible way to weave landscape architectural history into the matrix of modern challenges. Beautifully done

[Jim Scott @jimscott]

I have first hand proof the disintermediation is real. I have modernized three enterprise applications in geospatial technology in three weeks. These complex apps took 3years to build and one of them had the benefit of a design sprint - reducing time from concept to prod from 3years to five days for design - most orgs have no idea what is happening right now.

[Jim Scott @jimscott]

Love it. The guy being interviewed literally disassembled the whole premise. As of four weeks ago there is no moat. PLTR has FDEs Why? Because that's where the $ decisions are made - no FDE, no supplemental services contract - that's where the deeper capture lives. Follow the procurement officer

[Jim Scott @jimscott]

Same. What's also extraordinary is the capability to build a shared knowledge graph that can then be organized on a Wardley Map and you can define the edges and area for optimizing human input. A responsive and fluid framework focuses the application of energy and creativity. Compounding is the next theme. The limits to growth are giving way

[Jim Scott @jimscott]

Hey Deep, unless you can explain your process and demonstrate its capacity for empathy at a real human connection, you are just another algorithmic error in understanding the values of community and the needs of individuals. These processors have been around since the 1980s. Faster and richer graphics does not a more generative environment make.

[Jim Scott @jimscott]

There’s an entire universe of people with deep domain knowledge and strong vision who lacked one specific skill: programming.

Designers, planners, managers, analysts — they’ve always conceived what needed to be built. Now they can actually build it.

The bottleneck was never the thinking. It was translating that thinking into code.

AI removes that bottleneck. That’s not the death of software engineering — it’s the rise of design as the primary process for creation.

[Jim Scott @jimscott]

@rohanpaul_ai @ssankar We are reaping the narrow signals from all the STEM noise generated over the last 15 years. Unfortunate. We need more people who can transcend boundaries with their thinking.

[Jim Scott @jimscott]

This is so funny. Its called the design process. Too bad its only taught in specialized education settings. Define the problem, generate a hypothesis, collect data, synthesize and generate a conceptual solution. Challenge the assumptions, test the solution, iterate and refine to get to a design development stage, generate detailed design and specifications, plan the implementation, Construct, Monitor, review and summarize. Implementing this with agents is intuitive and the hard part is dealing with zero lag effects.

[Jim Scott @jimscott]

@bcherny Seems like a waste of energy, a salve for designers to retain their grip instead of coaching their agents to expand their creative range

[Jim Scott @jimscott]

@kyle_e_walker Yes. And Peter-OpenClaw just said we will not need the apps anymore. No wysiwyg. We just need trusted data agents and imo, good critical thinking skills. The intel will self-express for our preference. We need to be able to ask: "what is the essence of the problem?"

[Jim Scott @jimscott]

@MashfordMahute There is no Free geospatial lunch

[Jim Scott @jimscott]

@googleearth Hey, GE. Let's go way far away from viz. let's peel away the gl-candy and get into the geo-physics, the Terran and surface analysis, and let us explore the texture, diversity, and value-based descriptors of the 3D environment. How about it ?

[Jim Scott @jimscott]

@a16z He does not know

[Jim Scott @jimscott]

There's generaliam and then applied generalism. Enroll in a design school/college.

You get: studio environment, structured, critical thinking, problem solving, problem discovery and synthesis to conceptualize solutions. You learn how to transition from concept to reality eg. building specs.

There is a focus on collborative engagement, presentation and pitching/defending one's work and being able to look at things from an audience/user perspective.

[Jim Scott @jimscott]

The human is exposed as the weak link in this transformation. The first recommendation should be a context for the originator to guide all the downstream plans and actions. Claude is ready, willing and able to formulate generative productivity. It requires the human to learn how to shed specialist thinking and grasp the potential of collaborative engagement.

[Omar Sanseviero @osanseviero]

Today, you have the most powerful exoskeleton for the mind ever created at your fingertips.

You can kick off your day writing, spend the afternoon as an engineer, and go to bed as a designer. You can dive into the complexities of physics and use that knowledge to build a new software system.

We are moving from hyper-specialization to a world of generalists, where roles matter less and ideas matter more. The barrier to entry has collapsed.

Follow your curiosity wherever it leads. The world doesn't need more people who do one thing perfectly. It needs people who can bring ideas to life.

Build the future.

[Evan Applegate @youwillmakemaps]

She has a gallery show through Nov 30 at the Domaine du Rayol in Var, France

[Recall @recallnet]

How is AI rewiring the human brain? Recall Labs CEO @andrewxhill dives into this and more in the latest Thinking on Paper podcast.

🎙️ https://t.co/jZjyDZBbLx

[Jim Scott @jimscott]

Please, the regulations, zoning and other land development codes are not the problem. It’s the bureaucracy and lack of sophistication by leveraged builders who expend too much effort going for the minimal adherence and fail to commit to responsible practices. Just look at how many developments avoid or use minimal floodplain modeling and exceed reasonable densities by cramming 40’ lots into dead street grids. They fail to incorporate environmental design principles and then sink too much capital into over engineered infrastructure. They are stuck in the 1980s.

[Jim Scott @jimscott]

No, there are none. The crux is that no one has executed spatial awareness into a parallel model to LLM. We need a LSM-Large Spatial Model. It will probably not arrive from the mapping community. Current LLMs are pretty good at spatial inference from text patterns. They cannot think spatially however. There will be incremental steps. Constrain geographic extent and build spatial relationships using topological connections, aggregate data into relational grids like H3. Build agents that can generate these two logic on-the-fly with two dimensions, create a 3D model, Scale multiple 3D models to ‘overlay’ geographies in realtime. The challenge is life now with FSD and directional navigation at Tesla, Waymo etc. We have a long way to go.