Sign in

Fred Hebert

@ferd.ca
2.1K followers 190 following 560 posts

Principal SRE @ honeycomb.io, Tech Book Author, Resilience in Software Foundation board member, Erlang Ecosystem Foundation co-founder, Resilience Engineering fan. SRE-not-sorry. blog: ferd.ca notes: ferd.ca/notes

PostsRepliesMedia
Fred Hebert @ferd.ca · 27/09/2026
In 1997, James Reason published 'Managing Organizational Accidents'. It was influential in popularizing his concepts around the Swiss Cheese model, his specific brand/recipe of safety (and just) culture, etc. and discussed ever since. But it also has this line in it just throwing incredible shade:
Text from a book:

For example, some individuals are considerably more absentminded than others. If the person in question has a previous history of unsafe acts, it does not necessarily bear upon the culpability of the error committed on this particular occasion, but it does indicate the necessity for corrective training or even career counselling along the lines of 'Don't you think you would be doing everyone a favour if you considered taking on some other job within the company?'. This is the way that management acquires some of its most distinguished members. Absentmindedness has nothing whatsoever to do with ability or intelligence, but it is not a particularly helpful trait in a pilot or control room operator.
[...]

"This is the way management acquires some of its most distinguished members" is manually highlighted and a '!!' is in the margin.
23111
Fred Hebert @ferd.ca · 17/08/2026
this seems more accurate given the frequency of errors right now
the github unicorn from its error pages, without the horn.
0121
Fred Hebert @ferd.ca · 14/08/2026
This is the "abstraction-decomposition space". It maps a troubleshooting workflow from a system symptom ("no communication to tape or disk") via a diagonal walk to isolate a faulty component ("short-circuited transistor"). This is such an incredible visualization of so many complex elements!
A grid diagram plotting an Abstraction–Decomposition space. The vertical axis, Abstraction, runs from Physical (bottom) to Purposive (top) through five levels: Physical Form, Physical Function, Generalized Function, Abstract Function, and Functional Meaning/Purpose. The horizontal axis, Decomposition, runs from low resolution/coarse (left) to high resolution/fine-grained (right) through five columns: Whole System, Subsystem, Functional Unit, Circuit-stage, and Component. Fifteen numbered nodes connected by arrows trace a diagnostic troubleshooting path through the space, ending at a burnt-looking transistor short-circuit.
0124
Fred Hebert @ferd.ca · 14/08/2026
The 2nd one (same paper) shows aggregation, where a nuclear power plant can be described as a whole conceptual system (top), as functional sub-units with material elements (middle), or as granular units that get closer to the material world (bottom).
Figure 6. Three levels of representing a power plant's functional properties as flow structures, connected by large upward arrows indicating increasing abstraction. The top level shows overall energy conversion; the middle level shows mass and energy flow subsystems; the bottom level shows detailed component flows through the boiler, air/gas, and feedwater paths.
162
Fred Hebert @ferd.ca · 14/08/2026
The first one is the abstraction hierarchy from Coping With Complexity (1981). This shows how requirements and design tends to go top-down based on goals that must be met by the system. However, signals of problems tend to go bottom-up, from a local issue showing up as disrupted functionality.
Figure 4. The abstraction hierarchy used for representation of functional properties of a technical system.

The levels, listed top to bottom, are:

Functional Purpose — production flow models, control system objectives, etc.

Abstract Function — causal structure; mass, energy, and information flow topology, etc.

Generalised Functions — 'standard' functions and processes, control loops, heat transfer, etc.

Physical Functions — electrical, mechanical, and chemical processes of components and equipment.

Physical Form — physical appearance and anatomy, material and form, locations, etc.

To the right, a large diagonal line divides two opposing arrows. A downward arrow on the right labeled 'REASONS FOR PROPER FUNCTION,' 'REQUIREMENTS,' and 'PURPOSE BASIS'. An upward arrow on the left labeled 'PHYSICAL BASIS,' 'CAPABILITIES,' and 'CAUSES OF MALFUNCTION'.
1102
Fred Hebert @ferd.ca · 10/07/2026
at this point why not just host my repos on the staging servers, github
a search box from the github UI showing the word 'flag' typed in, and then a tooltip displaying '*flag* will be ignored since log searching is not yet available'
081
Fred Hebert @ferd.ca · 03/07/2026
These Angine de Poitrine fans got the best flags I’ve seen in a while; that’s some amazing concert gear.
Angine de poitrine at La Noce 2026 on stage, with two fans in the crowd waving large Quebec flags where the blue squares are replaced by black with white polka dots, the fleur de lys are golden, and the white is covered with black polka dots to match the band’s aesthetic.
1272
Fred Hebert @ferd.ca · 19/06/2026
I am still awaiting my anthropic settlement money
Strength of 597 / top 7% with Erlang/elixir author descriptionParody of B. Kliban’s cartoon where a police officer kicks a pedestrian to the side making way to a short man admired by women and pets, saying “out of the way swine, a cartoonist is coming!” Except the words cartoonist is replaced by “Erlanger”
1150
Fred Hebert @ferd.ca · 03/04/2026
Every time I am faced with this Swedish login form, I have to say 'Logga in' out loud with a terminator voice. It is one of the few rules we can't change and must be respected.
swedish login from with a button that says 'Logga in'
2271
Fred Hebert @ferd.ca · 31/03/2026
I don’t know that it’s fully invalid as a construct or been strengthened enough over time, but I liked this little distinction by Flach (1995) to show the risk of turning it from a description of a phenomenon to a causal process, and then Hollnagel’s later complaints that sort of crystallizes it.
Underwood distinguishes five levels of con-cepts. However, the distinction between Level 2 and Level 3 is where I would like to focus:
A Level-2 concept is one which summarizes the operations used to define a phenomenon and therefore merely identifies the phenomenon. I call it phenomenon identification or phenomenon naming. The definition of the phenomenonimplies not one thing about a causal process or condition over and above the operations per se.
(P. 198)
Level-3 concepts name or identify a phenomenon just as do Level-2 concepts; but, the name is applied to a hypothetical process, state, or capacity as a cause for the observations indicating the phenomenon. (P. 200)
Is SA a Level 2 concept or a Level 3 concept?
And why should anybody (except a fuzzy academic such as myself) care? Underwood's answer is that Level 2 and Level 3 concepts are based on the same formal operations but are
"'thought about' differently by psychologists"
(p. 202). In particular, he warns that "occasionally, having defined the term at Level 2, the writer may slip and talk as if it (the defined phe-nomenon) is now causing itself" (p. 202).
Figure 1, which is reprinted from Underwood
(p. 203), illustrates the difference between these two levels of concepts. In describing this figure,
Underwood wrote,
For the Level-3 concept I have drawn bidirectional arrows between X and Rd. For in using these concepts the investigator infers a state or process only if a reliable difference in response occurs and then says that this difference is caused by the state or process (X). Differences in X are in turn caused by Sm. If this sounds to you like scientific double-talk, then at this point I must agree. And it should be mentioned that Level-3 definitions do not always make circularity of the inference so obvious as I have made it here, but it is inevitably present.
(P. 203)
Consider some definitions for SA in light of Underwood's cautions.

Figure 1. A comparison of Level 2 and Level 3 concepts.
Sm indicates stimulus manipulation (i.e., independent variable). Rd indicates response differences (i.e., dependent variable). A Level-2 concept is defined by referring directly to the relation between Sm and Rd. A Level 3 concept identifies a state (X) as causing Rd and this state is, in tum, related to Sm.Folk models typically describe measurements that reflect an important aspect of the human condition but that refer to intervening variables - representing intermediate mental states - rather than to observable performance. As shown by Figure 3.2, performance can be characterised in many different ways, such as degree of efficiency, speed and precision of actions, performance deviations or 'errors, etc. Performance depends on individual attributes such as experience, motivation, capacity, skills, etc.
Performance is also affected by working conditions, which in turn can be described by means of a number of general and specific factors. Folk models propose that one or more of the intervening variables are strongly correlated with the quality of performance, and that measurements of this intervening variable provide essential information about how well the individual is able to accomplish a given task. Some of the more common intervening variables are shown in Figure 3.2.

In order for a folk model to be of practical use, it is necessary to assume that the measurement - or rather the hypothetical state that the measurement refers to - is affected by the working conditions in the same way that an individual being in the situation would be. (It also implies that the hypothetical state is the prime determiner of operator performance.) It must also be assumed that the measurement can be used to predict short-term changes in performance, since that is the main reason for not just observing what happens.

Figure 3.2: Loosely defined measurements.

- Intervening variables: workload, situation awareness, attention, fatigue, WM management
- Working conditions: general / specific factors
- individual elements: Experience, motivation, capacity, skills,...
- Direct measures of performance quality: 
efficiency, speed & precision, performance deviations
130
Fred Hebert @ferd.ca · 23/03/2026
hell yeah @resilienceinsoftware.org swag is in! the law of requisite variety states that only variety in the regulator can destroy variety in the system being regulated. so if you need to deal with complexity you know you gotta join the club & begrudgingly increase complexity to keep things simple
Black hoodie showing 'anti complexity complexity club' in a red explosion based on the logo of the Resilience in Software Foundation.
1152
Fred Hebert @ferd.ca · 13/03/2026
Here's a quote I always loved from Gill Kernick's "Catastrophe and Systemic Change" that I feel you might like given these recent experiences. Even if it's hard to apply, it's been stuck in my mind since then. It sounds like this approach, or something close to it, is working well for you :)
I recall Jim saying, on a day when I felt particularly helpless: "The doors of the people who should be talking to you will likely be closed, stop knocking on them. Look for the open doors' This took me out of linear thinking, into the emergent maze of change and complexity. I've stopped knocking on closed doors. I've stopped engaging in activities or conversations that will do nothing but maintain the status quo. I've learned that systemic change requires disruption - that I too need to shift how I think and act. I've come to see that kindness can be more disruptive than aggression, that compas- sion can be more impactful than taking positions. I've learned to plant seeds, and not to worry about which ones live and which ones die.
That is the nature of complex change.
1120
Fred Hebert @ferd.ca · 08/03/2026
As usual @grimalkina.bsky.social is worth listening to. These principles are also worth considering and applying in all sorts of contexts. Here’s a sample from safety research I happened to read just yesterday (Dekker - Reconstructing human contributions to accidents, 2002) that aligns with it!
What is striking about many accidents in complex systems is that people were doing exactly the sorts of things they would usually be doing the things that usually lead to success and safety. Mishaps are more typically the result of everyday influences on everyday decision making than they are isolated cases of erratic individuals behaving unrepresentatively.
People are doing what makes sense given the situational indications, operational pressures, and organizational norms existing at the time. Accidents are seldom preceded by bizarre behavior. People's errors and mistakes (such as there are in any objective sense) are systematically coupled to their circumstances and tools and tasks. [...] What people do makes sense to them at the time—it has to, otherwise they would not do it. People do not come to work to do a bad job; they are not out to crash cars or airplanes or ground ships. The local rationality principle says that people do things that are reasonable, or rational, based on their limited knowledge, goals, and understanding of the situation and their limited resources at the time. […] failures are baked into the nature of people's work and organization; that they are symptoms of deeper trouble or by-products of systemic brittleness in the way business is done. It means having to find out what people did back there and then actually make sense of it given the organization and operation that surrounded them.
To explain outcome failure, it is necessary to convert the search for human failures into a search for human sensemaking. The question is not "where did people go wrong?" but "why did this assessment or action make sense to them at the time?" Such real insight is derived not from judging people from the position of retrospective outsider, but from seeing the world through the eyes of the protagonists at the time. When looking at the sequence of events from this perspective, a very different story often struggles into view.
2295
Fred Hebert @ferd.ca · 17/02/2026
This sent me skimming through Accelerated Expertise by Robert H. Hoffman et al. again. I was wondering if there's a backwards way of going "if expertise transfer works, domains have to be similar", but they make the point that Transfer is incredibly difficult to predict and measure :/
Table 5.2: Example of transfer surprises. 3 columns are given: transfer tasks, expectations based on common elements theory, actual outcomes.

Example row: Rodeo riding vs. tournament jousting; expected positive transfers due to similarity in riding and falling without injury; found a large negative transfer because rodeo riders look to the back which doesn't work in jousting.
120
Fred Hebert @ferd.ca · 17/02/2026
here, I tried it and did it in ~12 minutes, fully vectorized and editable with an off-the-shelf app like draw.io. Now it's fully maintainable and styles modifiable, without working directly in SVG, and it's all costing nothing extra.
A manual remake of "A successful Git branching model" by Vincent Driessen (nvie.com). Flowchart illustrating a Git branching strategy with time flowing downward (bold arrow labelled "Time" on the left side). Five vertical lanes across the top: "feature branches", "develop" (bold), "release branches", "hotfixes", and "main" (bold).

Coloured circles represent commits on each branch: pink for feature branches, yellow for develop, green for release branches, red for hotfixes, and cyan for main. Arrows between commits show the direction of merges.

Text:

- "Tag 0.1" (on main, at the top)
- "Major feature for next release" (callout pointing to a feature branch merging into develop)
- "Feature for future release" (callout pointing to a feature branch on the far left)
- "Severe bug fixed for production: hotfix 0.2" (callout pointing to a hotfix branch forking from main)
- "Incorporate bugfix in develop" (callout pointing to the hotfix merging into develop)
- "Tag 0.2" (on main, after the hotfix merge)
- "Start of release branch for 1.0" (callout pointing to a release branch forking from develop)
- "From this point on, 'next release' means the release *after* 1.0" (callout with dashed line pointing to develop, where "after" is italicised)
- "Only bugfixes!" (callout pointing to commits on the release branch)
- "Bugfixes from **rel. branch** may be continuously merged back into **develop**" (callout with arrows showing merges from the release branch back into develop; "rel. branch" and "develop" are bold)
- "Tag 1.0" (on main, after the release branch merges in)

The diagram shows the complete lifecycle: feature branches fork from and merge back into develop; a release branch forks from develop when ready, receives only bugfixes, and merges into both main (creating a tagged release) and back into develop; hotfix branches fork from main for urgent production fixes and merge into both main (creating a tagged patch release) and develop.same as previous but in hand-drawn style
3210
Fred Hebert @ferd.ca · 17/02/2026
While LLM burnout isn't quite literal burnout, it's worth remembering that burnout is not purely individual. It's better framed as a workplace issue, due to stressors, mismatches between people and their jobs. "Take time off" is not very effective. www.youtube.com/watch?v=jVTf... is a good watch.
MANAGEMENT MANTRAS FOR MISMATCHES

• Burnout begins as a management issue.
  • "If you can't take the heat, get out of the kitchen" 
  • "You are going to have to do more with less"
• But if chronic job stressors are not managed successfully, the resulting mismatches between people and their jobs can lead to negative work outcomes (poor performance, absenteeism, turnover) and negative health problems (chronic illness, anxiety, depression).
• Chronic mismatches are often called "the pebbles in your shoeFIX THE JOB, NOT JUST THE PERSON

• Common response to a crisis is to help employees cope with ongoing stressors
• But burnout research points to a different response: 
  • HELP THE WORKPLACE MODIFY ITS SOURCES OF STRESS
• Given the current changes in the world, there are new opportunities to "think out of the box" and to experiment with new procedures and alternative strategies 
• SHARED RESPONSIBILITY FOR ANY SOLUTIONS (e.g. hybrid schedules, redesign workflow, etc.)
120
Fred Hebert @ferd.ca · 07/02/2026
Reading a text on Cognitive Systems Engineering (CSE) & its morality. It starts with the Ea Nasir tablet but then starts going hard and just won't let up, aiming for a morally relativist and nihilistic conclusion of "the real issue is our fake ass sense of absolute morality" The hell is this ride?
Engineering the Morality of Cognitive Systems and CSE; Figure 5.1 showing the Ea Nasir tablet with its full translation:

> Tell Ea-nasir: Nanni sends the following message. "When you came, you said to me as follows: 'I will give Gimil-Sin (when he comes) fine quality copper ingots.' You left then, but you did not do what you promised me. You put ingots that were not good before my messenger (Sit-Sin) and said: 'If you want to take them, take them; if you do not want to take them, go away!' What do you take me for, that you treat somebody like me with such contempt? I have sent as messengers gentlemen like ourselves to collect the bag with my money (deposited with you but you have treated me with contempt by sending them back to me empty-handed several times, and that through enemy territory. Is there anyone among the merchants who trade with Telmun who has treated me in this way? You alone treat my messenger with contempt! On account of that one (trifling) mina of silver which I owe?) you, you feel free to speak in such a way, while I have given to the palace on your behalf 1080 pounds of copper, and umi-abum has likewise given 1080 pounds of copper, apart from what we both have had written on a sealed tablet to be kept in the temple of Samas. How have you treated me for that copper? You have withheld my money bag from me in enemy territory; it is now up to you to restore (my money) to me in full. Take cognizance that (from now on) I will not accept here any copper from you that is not of fine quality. I shall (from now on) select and take the ingots individually in my own yard, and I shall exercise against you my right of rejection because you have treated me with contempt."As computational capacities have increased, the residual functions remaining to the human operator have been largely a litany of those tasks that computers had yet to be engineered so as to be able to accomplish. This left human beings, in Kantowitz's most evocative phrase, as the sub-systems of last resort. The moral concern with respect to a human operator's choice in terms of their own personal expectations, aspirations, dignity, and enjoyment of work was largely neglected, ignored, or deferred to the technical discussion of evolving automation capacities. Human hopes, desires, or more generally the hedonomic dimensions of work have very rarely been emphasized as imperatives, where the evolving technical capacities of the computer system have almost without exception served to drive the debate.Up until the emergence of CSE, many "systems-based" errors were thought of as the proximal problem of the closest human operator who was in some fashion error- prone or in other ways limited, biased, or even worse somehow "deserving" of such a fate (Arbous and Kerrich 1951; Dekker 2007b; Flach and Hoffman 2003; Wilder 1927). These have been couched in psychological terms such as lacks of attention, failures of situation awareness, and so on. It might be argued that these are somewhat more benign labels than one tellingly derived from the English military for pilot error: that is, a lack of moral fiber. These ways of stating the problem directly entail a stance that it is necessarily the advancements in technology that will circumvent such failures (Hancock 2000). Just occasionally, there arise assertions that collective social governances also bear some responsibility for such mishaps.

Blame, the moral pronouncement and disapprobation of the powerful collective on the behavior of the poor or poor individual or indeed any other less powerful collective, remains an intrinsic result of this divide. Here, error (sin, and its correlate-evil) becomes an inherent property or characteristic of that identified individual or minority. The excision of that one deviant operator or small group of operators quickly follows and thus the standard narrative is sustained. This notion echoes the disease model of harm in getting rid of the defect or one individual source. One great contribution of CSE has always been to look to expose this myth and to seek an alternative perspective...
Our western morality here emanates from an essential schizophrenia that derived from the concept of a munificent but omnipotent deity. [...] [Law as Western moralisty] has a vested interest in the status quo and so no great concern for changing it. Thus, CSE may be embraced by science and some limited segments of the business community but has yet to exert any proportionate social and moral impact because of the intransigence of the identified inertial institutions. [...]

5.9 CONCLUSIONS
As a result of the historical antecedents, which in the western world have divided process from purpose, our present circumstances generate complex work systems of great technological sophistication but erected upon ethical bases that are implicit, inconsistent, incomplete, and ultimately inoperable. Utilitarian, patchwork moral principles, derived from a dissolving base of religious dogma, are now morphing into ad hoc, secular legal arcana whose appeal to various poorly defined and developed human delusions, such as the notion of error, free will, and the omnivorous drive for greater efficiency, continues to subserve the dominant segment of profit-based capitalism (and polemically, see Dawkins 2009). [...]

Moral relatively has to be recognized, yet contemporary pragmatism employs and emphasizes an "as if" policy such that the search for moral certitude continues. Like error itself, morality is a convenient illusion that helps sustain societies in their necessary collective delusion. Morality is an intimate part of the narrative that we have always told ourselves (Homer 735 BC). [...]. Morality expresses social approbation and disapprobation and CSE subserves these moral goals (as well as other forms of purpose). What is good or bad, correct or incorrect are collective interpretations (most often after the fact) that support the individual and collective narrative that we have chosen as our conduit to collective reality (and, of course, see Nietzsche 1886).
171
Fred Hebert @ferd.ca · 03/02/2026
I've chatted with our AI folks at work around better support for investigations (rather than finding answers), arxiv.org/abs/2302.12389 has been real interesting in shifting framing away from "explainable answers" to "supporting decisions" in a way that lines up better with joint cognitive systems.
Table 3: The different ways that Explainable Al and Evaluative Al align against the human abductive reasoning process, using medical diagnosis as an example.

Event to explain is AI diagnosis in explainable AI, and Medical symptoms in evaluative AI; Hypothese are AI reasonins vs. Medical Conditions; Evidence are Models and explanatory info vs. Evidence for/against hypotheses; Judgment is causes of inputs vs. likelihood of conditionTable 2: Criteria for good decision support, based on the 10 'cardinal decision issues' [Yates and Potworowski, 2012] 

Options:  Help to identify options, as well as help to narrow down the list of feasible or realistic options
Possibilities:  Help to to identify possible outcomes for each of the identified options 
Judgement: Help to judge which outcomes are most likely 
Value:  Help to identify the positive and negative impacts on stakeholders for each of the identified options 
Trade-offs: Help to make trade-offs on the above criteria for each options
Understandable: Help to understand how and why the tools works as it does, and when it fails
130
Fred Hebert @ferd.ca · 13/10/2025
Last year I had grown some decently sized carrots, which felt pretty cool. This year’s harvest is just full on monstrously large eldritch taproots. They taste great, they’re just comically impractical to store.
17 large carrots, some over 12”/30cm long, most of them having 2,3, or more than 4 parts, all laid on the soil in a raised bed.
3250
Fred Hebert @ferd.ca · 19/09/2025
New SRE team swag is in
A t-shirt showing a pile of tires on fire, saying in large typeface: “fight fire with tire”, with a small byline below saying “SRE: fixing software with more software since January 1st 1970”

There are also 3 round stickers:

1. A bus labelled “SRE” dumping shit off a bridge onto a boat (in reference to the Dave Matthew’s bus tour incident), framed with text saying “Shitposting Reliability Engineering”

2. A bee on fire framed with text saying “honeycomb on-call incident commander”

3. A computer on fire with multiple bees flying around framed with text saying “honeycomb on-call incident responder”
2318
Fred Hebert @ferd.ca · 05/08/2025
designing truthful bumper stickers
top line on a fiery background: HELL YEAH I DRIVE FAST

The following text is below it and shrinks in font sizes until it becomes unreadable:

Modern urban design and traffic planning inadvertently encourage higher speeds: highway speed limit is based on the 85th percentiles and their entire form is based on increasing the safety of speeding drivers compared to encouraging actual safe behaviors. The good per-mile incident rates of highways then fed into the road design of denser cities with mixed traffic. This in turn inadvertently encouraged speeding within cities and shrunk usage of non-motorized transport. We should also look into systemic contributors, and a long known safety pattern for drivers is to match the speed of those surrounding them to remove the relative difference. What we must re-think is a broader set of speed-enhancing incentives in shaping traffic to truly value safety. Even the choice of metric used to optimize for the last century is based on miles driven and structurally under-emphasized denser areas as less important and
0295
Fred Hebert @ferd.ca · 30/07/2025
I’ve sowed white sweetclover for the last two years in the strip of soil next to my house and they’ve never even managed to sprout there. Yet these fuckers are growing wild right outside my fence, taunting me.
A flowering white sweetclover plant in front of a wooden fence
0110
Fred Hebert @ferd.ca · 12/07/2025
just gotta load up on context, understand the problem and what's already done so you can write a good soluti—
081
Fred Hebert @ferd.ca · 18/06/2025
I bought a pen that ended up shaped to only be comfortable with the dynamic tripod grip, but I have always written with the lateral tripod method. So I’m doing the most logical thing and I decided to learn a whole new writing grip, messing with decades of muscle memory.
A diagram of the 4 pen/pencil holding methods: dynamic tripod, dynamic quadrupod, lateral tripod, lateral quadrupod
11362
Fred Hebert @ferd.ca · 02/06/2025
James Reason taking no prisoners in Managing the Risks of Organizational Accidents
This brings me to another important point: outcome bias. There is a strong inclination, particularly among children and lawyers, to believe that errors having dire consequences arise from particularly reprehensible mental processes. In other words, they presume symmetry between consequence and cause. My shaving cream slip left me blowing bubbles and feeling foolish, but it was essentially inconsequential. Dr X's errors were not. However, it's the circumstances rather than the prior mental processes that largely determine the consequences. I could, for example, switch on the toaster rather than the kettle. The result would be mild annoyance. But if I had made a comparable error on the flight deck of an aircraft or in the control room of a nuclear power plant, the result could have been catastrophic. Such a slip contributed to the Chernobyl disaster.
1125
Fred Hebert @ferd.ca · 29/05/2025
Just starting to read Dara Z. Strolovitch’s “When Bad Things Happen to Privileged People” and it’s obvious early on that it’s gonna rule
These tensions and claims are thrown into relief by putting constructivist understandings of crises into conversation with work that treats crises as independent variables in the ways I began to do above: On the one hand, the bad things that shape the everyday lived experiences of marginalized groups are not considered crises unless and until they affect dominant groups. On the other hand, however, once a crisis is "hailed into being," it can have real and far-reaching political and material reverberations, both for the lived conditions of marginalized groups as well as for the possibility of opening windows of opportunity to address long-standing inequities.In particular, through inductive examinations of a wide array of evidence, I begin by showing empirically that crisis neither is a neutral descriptor of empirically bad things nor has it been a constant in mainstream political discourse. Instead, crisis politics have a genesis and a distinct historical trajectory, and both their origins and their arc have significant implications for the politics of marginalization and marginalized groups. We might assume, for example, that bad things are more prone to being labeled crises the more people they affect, the more they rupture with the past, and the more amenable they are to government intervention, but this is not the case. Instead, I show that crisis is endogenous to and recursive within politics. That is, I show that crisis politics are not only productive of but also the products of political battles and decisions, and that having a bad thing "societalized" and recognized as a crisis within dominant institutions is itself a political goal and victory that often has very real material and policy consequences, ones that often have feedback effects of their own. As lain White and Gauri Nandedkar put it, "crises are truth claims: they are invoked, they define, and, in doing so, they privilege certain ideologies or policy 'solutions' over others" in ways that may lead to different policy interventions and different attributions of political responsibility.
0112
Fred Hebert @ferd.ca · 12/05/2025
Seeing some cool sights
Statue of Maréchal Joseph Joffre at Champs de Mars, all being repaired and fixed
020
Fred Hebert @ferd.ca · 09/04/2025
overhearing LLM programming chat and this is all I can think about
remake of the old comic strip telling the user to "write a map-reduce in Erlang" to query the DB and resulting in "Did you just tell me to go fuck myself?" and "I believe I did, Bob."

This version has the first character ask "How do I make the AI learn things?", with the response "It doesn't it grows stateless context".

The next panel has the character clarifying "ok, it doesn't. How do I make it remember?" to which the other responds "You have to use the LLM as its own MCP server!", which leads to an unchanged original panel ("... I believe I did, Bob")
0182
Fred Hebert @ferd.ca · 23/02/2025
The right side stuff is native plants. Here’s the previous year’s plan for rotations. Left topmost box is an experimental one, bottom 3 cycle every year.
Gardening plan for 2024
130
Fred Hebert @ferd.ca · 22/02/2025
Winter months are spent planning and shopping for seeds, for those I don’t just finish prepping from the previous years harvest (stored in the cold and dark until then)
Garden plan for raised beds
130
Fred Hebert @ferd.ca · 20/02/2025
I finally had time to get back to it and now my little file synchronization mechanism works and supports parallel updates! I imagine it will take me a while to do the remaining 4 menu items since I've just cover my main bit but welp. Also a massive code clean-up is needed.
040
Fred Hebert @ferd.ca · 28/01/2025
New slogan proposal for the SRE team at work: FIGHT FIRE WITH TIRE
A honeycomb lanyard over a card with a poorly aligned label maker text going “FIGHT FIRE WITH TIRE” over “SRE”
2260
Fred Hebert @ferd.ca · 11/01/2025
it's okay i patched the new feature in
a power outlet that's poorly wired through another power outlet, seen in a big gap through a hole in a styrofoam board
4494
Fred Hebert @ferd.ca · 06/12/2024
Human-in-the-loop is a kind of "we're doing this responsibly" blanket for automation & AI. Experiments show large outcome variations based on where in the loop the automation and the humans are, and what sort of control is given. A question that should follow is WHERE in the loop is the human?
From https://www.nature.com/articles/s41467-024-51725-8

A Standard of care scenario: Readers 1 and 2 see the same episode and opt to recall or not-recall, if they disagree Reader 3 arbitrates. B AI standalone scenario: all decisions are taken by the AI Reader without human intervention. C AI single-reader scenario: Reader 1 takes the final decision using AI Reader input. D AI reader-replacement: same as (A) but with AI Reader replacing Reader 2. E AI band-pass scenario: AI Reader screens out episodes before Readers 1 and 2. Episodes with high scores trigger the recall decision directly, and episodes with low scores trigger the no-recall decision directly. The other episodes continue to the usual reader system. F AI triage scenario: AI reader triages the episodes before Readers 1 and 2. Episodes with high scores continue to the usual system, and episodes with low scores go through the path with only 1 reader.

Scenarios D and E show more robust improvements, and F is mostly useful for triaging when overloaded. Others have worse outcomes than Standard of care.Angry goose meme. Panel 1 is the goose asking "what is the human in the loop expected to do, exactly?", second panel, the goose is chasing a person screaming "WHERE IN THE LOOP IS THE HUMAN, MOTHERF*CKER?"
093
Fred Hebert @ferd.ca · 04/12/2024
I've been messing around with ncurses to make an interactive terminal UI for my file syncing app, and I realized that I was inadvertently writing a custom text editor.
1230
Fred Hebert @ferd.ca · 03/12/2024
I haven’t opened it all year and the icon is doing this now
Duolingo app icon, but it’s somehow shattered like glass and also sweats?
110
Fred Hebert @ferd.ca · 22/11/2024
This week's paper is Jon Kleinberg and Manish Raghavan's Algorithmic Monoculture and Social Welfare (www.pnas.org/doi/10.1073/...) In this one, they show that better algorithms can lead to worse outcomes, particularly when they are adopted in a uniform manner. notes at ferd.ca/notes/paper-...
Regions for different equilibria. When human evaluators are more accurate than the algorithm, both firms decide to employ humans (HH). When the algorithm is significantly more accurate, both firms use the algorithm (AA). When the algorithm is slightly more accurate than human evaluators, two possible equilibria exist: 1) One firm uses the algorithm and the other employs a human (AH) or (2) both decide whether to use the algorithm with some probability p. The shaded portion of the green AA region depicts where social welfare is smaller at the AA equilibrium than it would be if both firms used human evaluators. The AA and HH areas are of almost equal size and cover most of the chart; a narrow sliver shows AH as an area; the shaded section within AA is almost the same size and right next to the AH sliver
071
Fred Hebert @ferd.ca · 18/11/2024
The SRE-not-sorry one also got turned into sew-on team badges with the emote we used when killing hosts or containers :)
Stabby bear emote (copyright unknown) as an embroidered badge, with the text stating "SRE-not-sorry"
47011
Fred Hebert @ferd.ca · 13/11/2024
made an alternative logo for people who like principles of anarchism but won't commit to the full ideology and will instead just apply small versions of them in every day life. it's called anarcho-lowercapitalism.
the traditional logo of anarchism (capital letter A in a circle) and an arrow pointing right to the @ symbol
1200
Fred Hebert @ferd.ca · 09/11/2024
me going back to bed as soon as the system is "stable"
A warehouse with what looks like fridge boxes stacked high. Two fridge boxes next to each other are held up high seemingly in the air, as a tower of three stacked fridge boxes, collapsing at a 45 degrees angle, press them into place as if it were an arch.
07716
Fred Hebert @ferd.ca · 31/10/2024
My costume at work for Halloween is OceanGate safety technician. Also I am on call right now.
Fred wearing a cap saying "OceanGate Security Technician" and a headset, in front of a greenscreen image showing the gaming controller-operated submarine in a 3D render
0300
Fred Hebert @ferd.ca · 24/10/2024
The t-shirt version just made it, it's even better. Welcome to further trade-offs and goal conflict decisions!
A white outline on a dark green t-shirt. The design is hand drawn. On the top right, Rasmussen's drift model showing the three boundaries (economic, workload, performance) with the marginal error boundary is displaying the operating point shifting towards failure.
130
Fred Hebert @ferd.ca · 14/10/2024
Canadian thanksgiving means carrot harvest and hell yeah what a year for a single 4’x4’ bed mixed with radishes, with some damn monsters in there
2 litres ice cream tub overflowing with Nantes carrots, with more on the ground. There’s a knife in the ground which shows a 3 inch scale. There’s a hand holding a large carrot about 11”/28cm in length by roughly 3”/8cm at the widest
1290
Fred Hebert @ferd.ca · 11/10/2024
New custom stickers are in. Welcome to trade-offs and goal conflict decisions!
A white outline and transparent sticker on a grey laptop. The design is hand drawn. On the top right, Rasmussen's drift model showing the three boundaries (economic, workload, performance) with the marginal error boundary is displaying the operating point shifting towards failure.

The rest of the sticker shows the classic trolley problem meme, except both rail paths lead to 5 people tied to the rails, with a sobbing operator having to choose between two identical failure paths.
2202
Fred Hebert @ferd.ca · 10/10/2024
Today’s #FancyChips: Calbee’s Wasabi-flavored chips. They taste like wasabi, as advertised on the bag.
Green bag, with pictures of chips and wasabi roots. There’s text in Japanese I can’t translate as well.
040
Fred Hebert @ferd.ca · 04/10/2024
My ISP put locks over the coaxial cables they installed, and it turns out that you don’t need to ask them nor buy any tools if you want to move shit. Snap a wooden coffee stirring stick, insert both halves to create pressure between the crimp sleeve and the connector, then screw out to unlock
A coaxial cable. One end of it has a crimp sleeve with two wooden sticks inserted, and the other end shows the hex head next to a removed crimp sleeve.
2110
Fred Hebert @ferd.ca · 28/09/2024
*Chef’s kiss*
HN thread: Role of Deliberate Practice in the Development of Creativity (utexas.edu)


Comment: 
For those who don't have the time or energy to read the whole thing:
I asked NotebookLM to generate a podcast conversation out of this PDF.
Enjoy!
<url>
0102
Fred Hebert @ferd.ca · 20/09/2024
opened a PR up today that had some absolutely hellish one-liners, and I figured it needed adequate commenting. it's unlikely to get merged, but this feels like the least one can do to apologize to people who'll have to review it.
a minimap view of visual studio code showing Woman in Despair by Michelina Sarao but in ascii art, left as a comment in some file
060
Fred Hebert @ferd.ca · 19/09/2024
Reading "Disastrous Decisions" and saw a neat reference to an (accidental?) social breaching experiment showing how production imperatives can implicitly prevent safety. With priorities strongly aligned, calling for a stoppage about a conflicting value likely demands a lot of courage from workers.
I was once asked by a mining company CEO to take safety culture soundings [...]. I was told I could stop mining in order to talk to workers if I wanted to. So, at one mine, I asked that the whole operation be stopped in order to talk to miners at the face. [...] On arrival back at the surface, I was greeted by the mine manager who [...] went on to tell me quite aggressively that the stoppage I had requested had cost $20,000 worth of production. I was shocked. If he was willing to speak in this way to me, a representative of the CEO, it would be a very brave miner who tried to stop production for safety reasons. I am reminded of the work of the sociologist, Harold Garfinkle (1967). He suggested that the be  He suggested that the best way to understand the rules that are implicitly operating in a social order is experimentally to disrupt them. The manager's reaction to my disruption of his social order demonstrated the power of the production imperative operating at his mine.
163
Fred Hebert @ferd.ca · 19/09/2024
This garbage is not only pushed down an ever growing set of products that never really needed it, you’re also continuously made to participate in their elaboration without ever being asked. I can only hope this collapses relatively soon.
Option menu from LinkedIn stating “Use my data for training content in AI models” with a button set to On
191