Author: paschmann

Revopoint Inspire 2 – 3D Scanner

paschmann

3D scanner has intrigued me for a couple of years. The lure of being able to quickly and easily recreate something from a physical scan seems like the holy grail of self sufficiency and reducing your reliance on manufacturing or availability. 1.) With a couple of bike related projects that I want to tackle, 2.) the prices of the devices falling easily with the reach of consumers and 3.) the reasonable accuracy you get thanks to the underlying technology – it seems like it was the right time to get started.

Firstly – the device has been awesome, doing exactly what I anticipated and expected. Albeit a slightly cumbersome and slow learning process to understand the nuances of these devices. It is not the technology but rather the workflows involved to turn a “simple” scan, into a “simple” digital representation. There are a variety of steps in different tools and software, and the reality is that, as always, the devil is in the details. read more

Race: 2025 Ironman 70.3 North Carolina

paschmann

4 years in a row! A pretty solid training block going into this race despite some last minute travel to Germany for work.

2025 NC HIM Goal Plan vs Actuals
– Swim: 26:00 (Actual: 25:55, 140th OA)
– T1: 5:00 (Actual: 4:30)
– Bike: 2:10 (235W, 245Nor AVG < 160BPM) (Actual: 2:19, 150HR, 234W Avg, 239WNor, 18th OA)
– T2: 2:00 (Actual: 1:30)
– Run: 1:29 (6:35 @ 168bpm) (Actual: 1:39, 8W, 159HR, 60th OA)
– Finish: 04:20:00
– Result: #3 AG, 20 OA, AG Winner 4:15?, OA Winner 3:59?
– Training Load:
– CTL: 75 → 98 (Peak) 99 (Race)
– Bike Load: 46 (Peak) 45 (Race)
– Run Load: 35 (Peak) 36 (Race)
– Biggest Week: 13.1 Hours, 864 TSS
– Recovery Week: ~9 hours
– Avg Week: ~12 hours

A Definition of AGI

paschmann

An interesting paper which outlines a model for quantifying the concept of AGI, which is useful in determining the advancement of AI solutions.

A Definition of AGI

The lack of a concrete definition for Artificial General Intelligence (AGI) obscures the gap between today’s specialized AI and human-level cognition. This paper introduces a quantifiable framework to address this, defining AGI as matching the cognitive versatility and proficiency of a well-educated adult. To operationalize this, we ground our methodology in Cattell-Horn-Carroll theory, the most empirically validated model of human cognition. The framework dissects general intelligence into ten core cognitive domains-including reasoning, memory, and perception-and adapts established human psychometric batteries to evaluate AI systems. Application of this framework reveals a highly “jagged” cognitive profile in contemporary models. While proficient in knowledge-intensive domains, current AI systems have critical deficits in foundational cognitive machinery, particularly long-term memory storage. The resulting AGI scores (e.g., GPT-4 at 27%, GPT-5 at 57%) concretely quantify both rapid progress and the substantial gap remaining before AGI.

The Peter Principle

paschmann

The Peter Principle is a management concept formulated by Dr. Laurence J. Peter in his 1969 book “The Peter Principle: Why Things Always Go Wrong.”

The principle states: “In a hierarchy, every employee tends to rise to their level of incompetence.”

Employees who perform well in their current role get promoted
They continue getting promoted as long as they perform well
Eventually, they reach a position where they’re no longer competent
Once incompetent, they stop getting promoted (they’ve “plateaued”)
They remain in that position, performing poorly

This means that over time, every position in a hierarchy tends to be filled by someone incompetent to do that job. Peter observed: “In time, every post tends to be occupied by an employee who is incompetent to carry out its duties.”
And therefore: “Work is accomplished by those employees who have not yet reached their level of incompetence.”

The key insight is that competence at one level doesn’t guarantee competence at the next level. For example: read more