Recent content by harborsparrow

This may seem off topic, but it might be related too, since we're talking about the phenomenon of software development managers possibly skimping on best practices to save money. Remember that Toyota fiasco where cars would accelerate no matter what the driver did? The very description of the...

I don't think Amazon really expected to make money from the Kindle hardware anyway. Where they make money is in sales of Kindle books for reading on their free tablet and phone reader, which IS fairly friendly albeit full of annoying ads.

Factchecker, we'll have to disagree about the usefulness of code reviews. I recall them catching many issues, but I also recall them becoming far less effective when the software engineering pinheads came up with the idea of basing performance reviews on how few faults were found in one's code...

OTOH, had Amazon not pushed forward into ebooks and created the free reader, my house would still be cluttered with hundreds of old sci-fi and fantasy books so that I could reread them someday. Instead, I have them on the Kindle software client provided by free by Amazon. Cleared out like an...

AI should not, in my view, be used in weapons at all. This will be an unpopular view. I feel certain it has been happening for many years already. In the 1980's at Bell Labs, I worked in software for long-haul lightwave communications and later for managing networks of these systems. I...

AI "assisting" programmers is indeed extremely helpful. I use it too. I think the enquiry here was whether AI agents can or will or should be trusted to do the whole development process from a specification or design fed to it. It is impossible to know. Just remember that the U.S. and Russia...

I agree that AI will always be faster than humans, and managers will use it to generate software if it seems to save money. But because of the probabilistic nature of AI's underlying algorithms, I think AI coders will always misunderstand a few important things that (if it is true) would make...

My comments were in the context of whether AI can be guaranteed to perform well. AI that produces code cannot be made more reliable just by making it (i.e., the code-generating AI agent) open source where a lot of eyes review it's source code, because AI is Not Just Code.

There is a lot of wisdom in many of the above comments. My two cents is, don't go overboard. Code documentation ought not be a design doc, though people may feel falsely reassured by the presence of a lot of verbiage. Code documentation exists to assist future code maintainers in...

AI is more than just code, so making it open source may not be the same as for other software. AI also has a knowledge base and must be trained, and if I understand correctly (I am out of my comfort zone here), varying the data trained on, or even the order in which the algorithm is fed data...

Philips Hue was the very first provider of smart LED bulbs that can be controlled wirelessly. Over the years, as Philips sold ever improved models of smartbulbs, it became very difficult to identify exactly which model one of their bulbs is. Once I've installed a smart bulb, there is some...

As for "many types of AI", I believe it just about all involves various machine learning algorithms based on Bayes Theorem, is probability based, must be trained, and is subject to statistical type errors. If the code being written by an AI system can afford to have errors in it, fine. But if...

Old person here who wrote code in a gazillion languages including close to the metal. Elias001 is ahead of the game by learning C and assembler to understand precisely how memory is being used in a program. However, as programs historically grew more complex and larger, people found that C and...

IMO, current AI systems are far from being able to take over the whole of software development. One might trust a chatbot to cough up specific working bits of code to plug in, or more likely, use as a template, thus making coding go faster. Code generators are useful for mundane formulaic...

Here is an interesting article about Elon Musk gambling on LLMs being the future: https://www.wsj.com/tech/elon-musk-xai-memphis-tennessee-power-dec4c70d?mod=djemCybersecruityPro&tpl=cs Note that LLMs run on server farms of PCs whose powerful graphics processors are used in clustered form to...