cowanon22

This isn't "statistical whataboutery", it is basic ethical reporting. The French media got it mostly right; The US media got it mostly wrong.

We must have a zero tolerance policy for abuse, but bad reporting and bad extrapolation serve no one.

I'm referring to the US media - they covered it very differently from the French media.

Should we also shut down schools, scouting programs, day cares, adoption, camping, etc? All of these have had almost exactly the same size and scope of scandal.

I don't disagree with you - the church has major problems some of which still need correction. But there is a clear media bias in the US - compare the American articles to a much more details one in French (Google Translate works well) https://www.lefigaro.fr/actualite-france/les-cinq-chiffres-c...

I'm not doubting major abuse occurred, and the church needs major reforms. But we should hold the numbers up to the same standard that we hold all other numbers to. There should be a detailed description of the sampling and data, a detailed description or the model used for the projection, and a detailed description and justification for any bias adjustments. Any results should have an error band. Otherwise, this is likely published with an agenda, not to find the truth.

First off, I'm not doubting that abuse occurred and it is terrible for the victims involved.

What sampling method did they use to arrive at this estimate? What is the error range and 95/99% confidence intervals? I haven't been able to find the original report online.

From some articles: The commission found evidence of 2,900 to 3,200 abusers – out of a total of 115,000 priests and other clerics – but said this was probably an underestimation.

We would really need to see what the method they used to arrive at the numbers - this smells like major hyperbole.

UPDATE: a better article is available in French: https://www.lefigaro.fr/actualite-france/les-cinq-chiffres-c...

The number they gave is cases over 70 years, which is a weird way of reporting it. Annually, it is about 2900-3000 priests per year. The article also mention that the current rate in about 1/4 of the rate in the 1970's, showing that some of the church reforms have actually worked. If you compare the French stats to what is being reported here, the anti-church media bias is flagrantly clear.

We don't know yet if there is anything special about human intelligence, or what the limitations of general intelligence might be. Are animal and plants a slower form of the same intelligence, or is there something qualitatively different? Can rocks and liquids be considered intelligent since it led to life, which eventually led to us?

Current AI/ML does not appear to have any properties of "life intelligence" - for example you can put an animal or even plant into an unfamiliar situation and it will often figure out a way to survive. AI/ML in the evaluation phase is often pretty dumb, and needs new training if anything changes. Reinforcement learning is probably the closest but still seem pretty limited.

I don't think the current approaches will lead to general intelligence. However, I do suspect that when the right theoretical breakthrough is made AI will rapidly become superhuman and humans will not be in control of what it does - it will simply iterate too quickly for us to compete in any way intellectually.

The truth is that artificial intelligence and machine learning today are still aspirational titles - the current tech simple does not learn or reason in any way that is similar to human or even biological reasoning.

Data mining and probabilistic pattern recognition are much more accurate descriptions, but don't sound as exciting.

It's definitely possible that true AI will one day exist, but it may be anywhere from 5 to 1000 years away. I suspect the current approaches will not resemble the final form when it comes to AI.

In my experience the most useful part of DDD is to have a common vocabulary for your projects - developers and end users should have some common terminology. For complex business domains it is good to have a glossary, and for your code to always use those words in the same way the business uses them.

Other than that, most DDD concepts are a bit dated, and really oriented around JAVA/C# in the early 2000s.

This is yet another paper where the title exaggerates the importance of the conclusion.

I found a copy of the original paper - there are several risks to the model that the authors mentioned:

- They only mutated conditionals to test coverage, not any of the other possible errors the test suite may be looking for.

- Equivalent mutations may be miscounted, particularly when developers test a lot of off by one errors.

- There data may not meet the assumptions of the Kendell r correlation used.

- Most of their data had low levels of coverage - previous research shows high coverage is needed before it is related to effectiveness.

- They did not account for object oriented code boilerplate code (getter/setters) which do not need to be tested, causing their counts to be potentially be off. This is major, as they were using only Java projects.

- They had very narrow inclusion criteria, so the results may not be generalizable to all codebases. The projects had to have over 1000 test written; the average LOC was generally in the 100k range.

This honestly sounds like some grad students final project, not advice for real world digestion.

Unfortunately the marketing has gotten ahead of the actual engineering - the features can be useful but it does take an adjustment period to learn its limitations.

I have Toyota Safety Sense 2.5 on a 2021 Camry:

Pre Collision system: pretty good, it did help once when a car suddenly pulled in front of me and rapidly slowed down. Pretty good at detecting pedestrians when backing up. A good safety net, but you should never rely on it.

Dynamic cruise control: pretty good, but you need to take control if traffic is totally stopped or a car suddenly pulls in our out in front of you.

Lane departure / lane tracing assist: almost useless - too many faded lanes and complicated patterns where I live.

Road sight assist: accurate but mostly useless

Blind spot detection: a good first pass, but you still need to look. It can miss motorcycles or bicycles on a slower road. As the article mentioned, it can also miss if something is about to go into your blind spot.

By far the most useful newer feature is the 360 view camera - it makes parking super easy.

I'm not expert is any of this, but what worries me is that political pressures, conflicts of interest, profit motives, rushed timelines, experts who have misled to get a reaction, and many studies with either short terms or small sample sizes are making it so that the average person has no idea what to believe.

"No Silver Bullet" is similar to "Moore's law" - an observation rather than an absolute rule. In both cases, the authors were only projecting into the future 10 years.

It is just measurement, not pessimism or optimism. The fact that "No Silver Bullet" is still true 35 years later show just how hard progress can be.

> Tor has done more harm than good for human rights

Sad but true - it seems like every technology that is made to promote free speech, anonymity, or non-centralization ultimately leads to bad actors using it and more authoritarian laws being passed. I'm not sure what the solution is.

> So, maybe age-ism is a thing, but I'd also argue a lot of age-ism can be reduced to "I never even tried keeping up and hyper-specialized myself into a dying corner even when the signs became clear I'm becoming outdated, so why doesn't anyone hire me?"

My experience has been exactly this. I've read many such complaints in the last 20 years of working, and in every single case the author has pigeonholed themselves into a long obsolete corner of the industry. There is a difference between 30 years of experience and 5 years repeated 6 times.

It is important to get onto the right team to have a long lived tech career. Your business domain should be big enough so you organically have to move with the rest of the industry. By 20+ years, you should be in a senior enough position to help lead the change.

I've never met a tech guru who was up to date and didn't retire or move on to other things on his/her own terms.

I agree with a lot of the points, however at most large companies solution architect is a significantly higher paygrade than developer. (I don't agree with this, but it is the case at every large place I've worked [non-FAANG Fortune 500].) Usually a career progression goes from developer to lead dev positions to solution architect, not the other way around. There is generally a big money drop if you go backwards.

This is why I now use tools like SonarQube (commercial static code analyzer) to drive most code quality reviews. SonarQube also gives justifications for all of it's recommendations, so it serves as a training tool as well.

The only I things I manually look for in reviews are overall readability, sensible naming, high level adherence to the architecture/design, and data system/database/API queries.

After 20 years of writing code and leading teams, I've found that most of the stuff developer blogs and the OOP community fret over simply doesn't matter in most cases. The high level interfaces are often more important than the low levels details, and static code analyzers do a decent job of pointing to problems. I've almost never seen bad code that didn't also have many issues in the static analyzer.

It also matters what kind of code you are writing - foundation code or something that will be used by thousands/millions of developers should have a higher bar than some internal one off. However, I've seen enough very profitable companies running on total crap code to believe that code quality is the most important thing.

Lol, spoken like a true corporate slave. Sad thing is anyone who works in McKinsey has more than enough talent to succeed without sacrificing their personal life. I guess the saying is true, McKinsey is filled will people who used to be interesting.

The paper does not specify message passing or reflection, both of which are essential to OOP. I think OOP lost its way as soon as C++ reinterpreted it in terms of only encapsulation, inheritance, and polymorphism.

According to Alan Kay (one of the creators of Smalltalk, the first OOP), message passing is more important to OOP than inheritance. If you listen to his old speeches, what he is describing sounds a lot more like microservices and VMs/containers than what most of the later languages turned it into. (He describes objects as "mini computers" that interact with each other using only public interfaces)

Lol, I've been using Windows for a long time and had no idea these shortcuts exist - I will probably use them all the time going forwards.

It seems unlikely that your co-workers have low IQ, but it does go to show that doing well on algorithms and toy puzzles doesn't necessarily lead to an interesting job. It seems odd that you would be assigned to a team that didn't personally interview you - is this normal practice at FAANG companies?

Your co-workers may just be zoned out and collecting a paycheck - at some point the workload can become so impossible that you just don't care anymore. This is especially true if you don't care about a company's mission, product, or customers. Unfortunately the tech boom has led to high turnover at some places, and people just jumping to the next big salary instead of picking a good long term fit.

The tech industry is big enough that you can do what makes you happy - the big name places sometimes aren't a great fit. I've found a pretty good career working at mid-tier Fortune 500 companies. There a lot of good opportunities all across the US if you're willing to forgo SV salaries, know your stuff (and can learn new tech quickly), and focus on helping the business instead of the latest tech fad. (FYI, I make 175k in the mid-west with 40hr work weeks and a 10 min commute, 5 weeks PTO, great healthcare)

With Spring this is actually true - there are better alternatives for Java, and there is too much magic under the covers. I use plain servlets, jsp and a handful of libraries for common operations, and it is so much easier to debug when something goes wrong.

Every quality job I've had consisted of 6-8 different interviews. However, all of those jobs also send me a schedule and list of job titles of who I will be interviewing with, and conducted all of them on the same day. At this point I consider it a big red flag if a company excessively spreads them out.

I'm fine with having 6-8 interview - every

I've found that these situations are actually quite useful from my side as a prospective employee. If the interviewers are treating this as a "gotcha" interview and it turns into a test instead of a discussion, then that's a pretty good sign that this job is not going to be a good fit for me. The best jobs I've had are where the interviewers treated me as a potential future colleague, and the technical questions were more of a starting point for a detailed discussion of different scenarios rather than a straight quiz.

Yes, most martial arts are more of cultural dance and ceremony than an actual effective fighting techniques. MMA/UFC has done more to advance the state of the art than a 1000 years of dojos ever did. From a cultural standpoint it's a shame that this martial art is lost, but I would be surprised if it had anything novel that would translate to real combat.

Personally I think we should stop using the words intelligence or learning to refer to any of these algorithms. It's really just data mining, matrix optimization, and utility functions. There's really no properties of learning or knowledge.

The report is short (8 pages) and is really just there to motivate the creation of a unified procedure for detailed collection of sightings; interestingly there appear to be 18 relatively recent incidents that cannot easily be explained and appear to have unknown propulsion characteristics. I wish they would release all of the images, videos, and data that they have, but it would probably reveal too many details of classified sensor systems.

Combined with the recent French report (and other government reports in the past) this does seems interesting - this is the first time governments have officially admitted that something unexplained is being seen regularly by experienced aviators.

I'm generally skeptical, but in this case I'm not too convinced by the debunkers either. However we would really need the full data to be released or available for review by multiple groups of experts to know anything.

> Like any other industry, 90% of medical “experts” are totally incompetent.

I would guess it's even worse, maybe 1-5% of doctors truly understand the underlying physiology and will holistically look at all of the information. Most doctors I've met use population heuristics and published guidelines, and if you have something rare they just throw stuff at the wall instead of trying to understand it.

The AI field has a history of over confidence; just look up Marvin Minsky's predications in the 1950's and 1960's. This had been repeated throughout AI history - a new approach is found, it has some promising initial results, and then progress gets stuck and the approach is basically halted for decades.

Machine learning and neural networks have prospered due to massive increases in computing speed and data, but I suspect it will also be a dead end in general AI. There will be some niche wins, some quite impressive, but the machine won't have the properties of an intelligent child or even a rat.

There seems to be a big difference in how actual neurons work - I think neural networks are misnamed. The brain is not a big matrix, biology does not appear to reinforce behavior using gradient operations. I think increased research into neuroscience and biomimicry is the most likely approach for general AI, but I think we are still far away and the current hype will just be one more dead end.