Awais Aftab provides a succinct explanation of autism, as per DSM-5, and how its diagnosis is based on a descriptive prototype rather than medical tests.

To be diagnosed with autism spectrum disorder according to DSM-5, a person must have ongoing difficulties in social communication and interaction in all three areas: trouble with back-and-forth social connection, problems with nonverbal communication like eye contact and body language, and difficulty making or keeping friendships. They also must show at least two types of repetitive or restricted behaviors, such as repetitive movements or phrases, needing things to stay the same, having very intense focused interests, or being unusually sensitive (or under-sensitive) to things like sounds, textures, or lights. These patterns must have been present since early childhood (even if they weren’t noticed until later when life got more complicated), lead to substantial impairment in functioning, and can’t simply be explained by intellectual disability (or other psychiatric disorders).

To “have” autism is simply to demonstrate this cluster of characteristics at the requisite level of severity and pervasiveness. It doesn’t mean that the person has a specific type of brain attribute or a specific set of genes that differentiates them from non-autistics. No such internal essence exists for the notion as currently conceptualized.

Autism spectrum is wide enough to have very different prototypes within it. On one end we have profound autism, representing someone with severe autistic traits who is completely dependent on others for care and has substantial intellectual disability or very limited language ability. At the other end, we have successful nerdy individuals with autistic traits and superior intelligence, often seen in science or academia, à la Sheldon Cooper. (Holden Thorp, editor-in-chief of the Science journals and former UNC chancellor, for example, has publicly disclosed his own autism diagnosis.) This wide range is confusing enough on its own, even without considering other conditions that can present with autism-like features.

Autism cannot be identified via medical “tests.” It is identified via clinical information in the form of history, observation, and interaction, and the less information available or the more unreliable the information provided is, the more uncertain we’ll be. To have autism is basically a judgment call that one is a good match to a descriptive prototype. We can get this judgment wrong, and we sometimes do get it wrong. (There is nothing wrong with this fallibility as such, as long as we recognize it. Lives have been built on foundations less sturdy.)

Addy Osmani shares his 21 lessons from working in Google for 14 years. This lesson stood out the most for me.

14. If you win every debate, you’re probably accumulating silent resistance.

I’ve learned to be suspicious of my own certainty. When I “win” too easily, something is usually wrong. People stop fighting you not because you’ve convinced them, but because they’ve given up trying – and they’ll express that disagreement in execution, not meetings.

Real alignment takes longer. You have to actually understand other perspectives, incorporate feedback, and sometimes change your mind publicly.

The short-term feeling of being right is worth much less than the long-term reality of building things with willing collaborators.

Shubhabrata ‘Shumi’ Marmar reflecting on the dramatic design of Mahindra BE6.

What you should know about dramatic designs is there is always been a long-standing debate about whether they’re good for a company to do or not.

Dramatic designs can draw attention; they can create aspiration, but they can also age very quickly, which is why a lot of companies are a little leery of doing super out-there designs for production cars.

You do that for concepts because the concept is about attention and not production.

Therefore, if you look at, for example, the Polo—the Polo didn’t age a lot, but the Polo didn’t also look super fresh when it arrived. It looked like a nice, clean, fresh design, and it looked like that for a long time because it wasn’t a super dramatic design.

Whereas every time Lamborghini, for example, to go right to the other end of this, when they have to build a new supercar, they have to really push a dramatic design out because it is a supercar.

It’s not going to get produced in large numbers, but it’s that drama that is part of that brand’s design story.

But when you take it to production, which will sell a large number of cars, the risk is if it’s too dramatic, then in two years it’ll look like the last trend rather than today’s fresh car.

That’s why dramatic designs—it’s a trick; it’s a good trick, but it may not always work.

Anand Sridharan talking about how predicting future events cannot help us predict the future consequences.

Herein lies the true unknowability of the messy world around us. It is not just we cannot predict major events. It is that, even if we could, we have no idea how their consequences will play out. It is impossible to reliably unravel a chain of future events in a manner that is useful in real-world decision making.

I used a 2-level thought experiment merely to illustrate that the world is unknowable at many levels. There is no reason to stop at two levels. Cause-and-effect plays out at many more levels, often with feedback loops tying certain consequences back to original events. In fact, in the real world, it is far from clear what the original causes are for whatever transpires around us.

This thread by Dror Poleg about how technology has the power to turn in-person work into scalable work.

In 1930, the union of American singers spent the equivalent of $10m on a campaign to stop people from listening to recorded music and watching movies with sound.

When films were silent, theatres employed local musicians to accompany each screening. But once films gained a soundtrack, local musicians were no longer necessary. The economic implications were significant: In 1927, around 24,000 musicians were employed in theatres across the US and Canada. But then came the first talking film — The Jazz Singer.

By 1930, some 7,200 musicians lost their jobs — 30% of the pre-talkie total. In some markets, such as New York and Cincinnati, musician unemployment reached 50-75%.

Over time, all theatre musicians were eliminated, and recorded soundtracks became par for the course.

The advent of records, radio, and talking films made creative work scalable: “300 musicians in Hollywood supply all the ‘music’ offered in thousands of theatres. Can such a tiny reservoir of talent nurture artistic progress?”

A hundred years ago, it seemed improbable that “canned music” would replace “real” music. Joseph N. Weber, president of the American Federation of Music, predicted that the public will not always accept “like-less, soulless, synthetic music.”

Edward More, the Chicago Herald Tribune music critic, agreed with Weber, stating that “the films have a long way to go before they can duplicate living musicians”

Films never managed to “duplicate” living musicians. They didn’t have to.

Disruptive technology doesn’t seek to “replicate.” More often, it sidesteps and makes old standards and processes redundant.

Records and talking films made music cheaper and accessible to a much larger audience. Most of the audience didn’t care about traditional quality.

As a result, we tend to underestimate technology’s power to turn in-person work into scalable work. In many “creative” professions, fewer people can already capture a larger share of the market than ever.

Such professions include programmers and designers, but also teachers and fitness instructors. A Peloton instructor earns about 12 times more than an offline competitor — and can service many more clients at the same time.

We assume that most professions cannot be scaled in the same way. But there is already evidence to the contrary. Many things that seem ridiculous to us now will seem obvious to our grandkids.

Thsi comment by TexanFeller on Hacker News explaining monotropism with a metaphor.

I find monotropism an apt way of understanding it. A normal person’s attention is like a flashlight they control that illuminates much of a room at once. Autistic brains are a tight beam flashlight, almost a laser for some, with its aim difficult to change. ADHD brains are more like a tight beam flashlight on a motorized mount that swivels in all directions, but you’re not always in control of where it swivels to…it’s like an AI constantly overrides your direction inputs and points the light at what it deems most exciting or urgent at the moment.

This comment by godelski on Hacker News about how folks in the long tail, who seem least productive, are the ones who end up changing the world.

You can’t have paradigm shifts by following the paradigm.

How I think of it is we need a distribution of people (shaped like a power law, not a normal).

Most people should be in the main body, doing what most people do. They’re probably the “most productive”.

Then you have people in the mid tail who innovate but it’s incremental and not very novel. They produce frequently (our current research paradigm optimizes for this). But there aren’t leaps and bounds. Critically it keeps pushing things forward, refining and improving.

But then there’s those in the long tail. They fail most of the time and are the “least productive”. Sometimes never doing anything of note their entire lives. But these are also the people that change the world in much bigger ways. And sometimes those that appeared to do nothing have their value found decades or centuries later.

Not everyone needs to be Newton/Leibniz. Not everyone should be. But that kind of work is critical to advancing our knowledge and wealth as a species. The problem is it is often indistinguishable from wasting time. But I’m willing to bet that the work of Newton alone has created more value to all of human civilization than every failed long tail person has cost us.

In any investment strategy you benefit from having high risk investments. Most lose you money but the ones that win reward you with much more than you lost. I’m not sure why this is so well known in the investment world but controversial in the research/academic/innovation world.

This reminds me of the book Slack by Tom DeMarco.

Dare Obasanjo talking about what it takes to be a leader.

Many people want to be managers to grow their careers but in reality being a leader is more important in the long run. A leader is someone whose authority is earned and people choose to follow as opposed to someone people have to follow because it’s their job.

Good leaders are…

• focused on the team instead of themselves. Good leaders want to develop others and help them grow. They are ambitious but it’s for the team’s success not just their own career.
• effective and transparent communicators. They also understand listening is part of communicating
• self aware about their strengths and weaknesses. Self awareness is coupled with a growth mindset. Weaknesses aren’t limitations but opportunities to learn.
• respectful to their team and act with personal integrity. They set positive examples of how people should be treated.
• effective at both short term execution and long term strategy. There is a vision for where we’re going and a path for how to get there.
• effective delegators. Good leaders encourage autonomy and avoid becoming decision bottlenecks while giving the team growth opportunities.

Robert Greiner reflecting on the recent acquisition of Bun by Anthropic.

Leaders don’t express their true beliefs in blog posts or conference quotes. They express them in hiring plans, acquisition targets, and compensation bands. If you want to understand what AI companies actually believe about engineering, follow the cap table, not the keynote.