jeyoor

The combination of high and climbing price to earnings ratios for a smaller subset of tech firms, outsize retail investment in tech (cloaked by people buying crypto), and macro environment factors like high interest rates stimulating risky lending has me swapping this bubble toward the top of the list.

See further: https://www.morningstar.com/news/marketwatch/20250123167/six...

There are Heterotopias in the US too. Some of the border regions in Southwest Texas intermittently occupied by cartels or smugglers come to mind.

I propose that by prioritizing profits and productivity we were already globally dependent on tools to find meaning in the world before generative AI. This seems to be the society we already have.

At first glance, these results appear fairly impressive as far as deep learning transfer studies go.

However, standard caveats about the limits of those approaches should still apply: e.g. https://spectrum.ieee.org/cars-that-think/transportation/sen...

In other words, I don't think MultiModel will be immune to fairly straightforward adversarial image attacks (although you might need a different adversarial network to generate them).

Furthermore, the problems being addressed by MultiModel (image recognition, natural language processing, machine translation) are problems that already have fairly robust deep learning results.

I'd be more interested if MultiModel showed significantly better results on problem areas that are currently difficult for standard deep learning approaches.

If you disagree with Taleb's politics, I recommend focusing on his scholarly publications.

http://www.fooledbyrandomness.com/FatTails.html

https://drive.google.com/file/d/0B8nhAlfIk3QIR1o1dnk5ZmRaaGs...

I've found the rigorous presentations of his ideas in statistics and finance to be very helpful.

In addition to the Postgres knowledge, this post also provided me with tremendous help in building a sane data architecture for RESTful web services in Haskell.

I agree! I found the list of potentially marketable hobbies great too.

It's amazing how much technological change we take for granted when choosing careers.

Kudos to this team for dogfooding their platform. That choice makes me far more likely to evaluate and eventually use it.

Anything like this available for .NET Core?

Which BaaS options are recommended now?

In some Sharepoint 2010 deployments I worked on, it was possible to create Sharepoint workflows that would would bog down, fail to process new entries, etc. once the list of items grew beyond 100-150 entries.

Admittedly, this was probably related to misconfiguration and database issues (i.e. having zero oversight or administrative maintenance of the underlying MS SQL Server). That specific local minimum might not apply to the context of the article (optimization in code and systems design).

>You could see all of this in a negative light. Run for the hills and disconnect! Reality or nothing!

>I have another idea:

>It's time to stop designing digital services to just be easy to use and start designing them to be understandable, accountable, trusted, and easy to use.

I think the author's triad of adjectives (understandable, accountable, and trusted) is widely applicable. It seems this article is more focused on software safety than political systems.

Following the author's analogy with the industrial revolution, I think the development of safety regulations for industrial workplaces in the US is fascinating. http://eh.net/encyclopedia/history-of-workplace-safety-in-th...

I wonder what safety regulations for software systems would look like.

I have tremendous respect for the decision to post the DIY guide alongside the product. I find this sort of kindness hugely inspiring.

It's good to see that privacy concerns are being voiced publicly. I'm not certain that class-action lawsuits are the most productive approach to take, though.

On the technical side, I was intrigued that the "black box" of multilayer neural networks could provide a legal defense for Facebook in this case.

Often, the fact that neural networks are harder to analyze compared to other AI techniques is cited as a detriment, but in this legal case it could prove useful.

I'm impressed with this project's messaging and support/feature pricing structure. I think trying to find more sustainable models for funding open-source software is interesting, and I don't recall seeing this exact approach used before (i.e. free version w/o LE or Graphana support). It'd be amazing if a percentage of each support contract gets donated back to upstream projects like icinga2, Graphana, and LE.

This article matches what I've been seeing anecdotally (especially at smaller tech firms and universities in the Midwest US).

I've been hearing more folks in research and industry express the importance of applying simpler techniques (like linear regression and decision trees) before reaching for the latest state-of-the-art approach.

See also this response to the author's tweet on the subject: https://twitter.com/anderssandberg/status/803311515717738496

I wonder if approaches like this will work for poetry (or prose) from different cultures. I'd be especially interested to see if some of these techniques are effective on languages markedly different from English. (Maybe Chinese and Arabic?)

On just a quick glance, the breadth of the topics covered here is stunning.

I also liked this quote from Lecture 23.

> A lot of times we ... confuse value with complexity.

> And many of the things that were the simplest in this subject are actually the most powerful.

> So be careful about confusing simplicity with triviality and thinking that something can't be important unless it's complicated and deeply mathematical.

The author seems to say that the no free lunch theorem (NFL) indicates that creating a "Universal Learner" (or artificial general intelligence) is an impossible task.

I disagree based on the following:

1. I think NFL's definition of a universal learner is broader than the definition used by the average AGI researcher.

In practice, we are interested algorithms producing behavior at near-human levels of intelligence, not absolutely universal learning algorithms.

2. NFL does not seem to directly address the overall effectiveness of applying a combination of algorithms based on real-world experience to everyday problems.

I think rigorous application of formal methods plus deployment techniques like unikernels have the potential to improve the quality of a lot of software.

It seems that markets such as India and China[1] are more comfortable with apps packed with numerous features (i.e. WeChat or LINE), whereas US apps seem to usually be more focused on minimalism [2]. Snapchat is a notable US counter-example.

[1] http://www.nytimes.com/2016/08/03/technology/china-mobile-te...

[2] https://blog.benroux.me/your-app-has-one-feature/

Things I enjoyed in this article:

1. The emphasis on the importance of general problem solving

2. The clarity concerning various slippery definitions. ("Scientist," "Big Data," etc)

3. The high-quality links and citations