If you're locked inside an ATM, have you tried banging on the case to alert passers by?

Almost nobody uses "untyped languages". Few of those even exist, with Forth and various assembly languages being the main examples. (C, with its type system that is as airtight as a sieve, gets an honorary mention.)

You're probably harping about dynamically typed languages. In such languages, the runtime still knows *exactly* what type every item of data has. These are not weakly typed. But what you obviously prefer are "statically typed" languages.

Static typing might statistically reduce some errors, but it certainly can't "eliminate whole classes". Consider "set_warhead_target(float latitude, float longitude)". Did the type system give you any protection from accidentally swapping the two parameters? That's really the problem that you're so worried about: accidentally using the wrong data value in the wrong place.

However, very few statically typed languages (with Rust being a notable exception) have eliminated the biggest source of type errors in computing: Null, which is a bogus placeholder that matches any pointer type (or reference type, depending on the language's nomenclature). So in many cases you have no less risk with static typing than you do with accidentally feeding a string into a Python sqrt() function. And in the case of C or C++, you can be much worse off, as in segfaults and remote exploits.

If you want a functioning society, everybody needs to be able to participate.

But I do not think people like you understand that. You just want to be inhumane and cruel to groups you perceive as subhuman. And that makes you society-destroyers. Not that any of you is smart enough to understand that. Or moral enough to see that being inhumane is maybe not a thing good people do.

I do not agree and there is evidence. For example, about half of larger code-samples generated by AI contains vulnerabilities. That is not "mostly great", that is "pretty bad". If these are systematic (i.e. somewhat predictable for attackers, who will also use AI), then this becomes "abysmally bad". A second result is that LLM use makes coders around 20% slower (!) on average, with most mistakenly believing they are faster. Cost of worse code are not included in that.

With the job-replacement, we will haver to see. Many jobs do not require much insight or skill. Those may get reduced. Or not. A key problem (besides hallucination and total lack of insight) is that the true cost of running and maintaining LLMs is still kept hidden. What we know does not look good at all. It may turn out that actual cost is, say, > $1000 per month per professional user. That will not cut it at all. It may also turn out the general LLMs become or already are impossible to update or train from scratch because there is too much AI slop already out there.

Mozilla is really broken.

Renault are doing fantastically well with their cars right now. They have won accolades from every corner for the new Renault 4, 5, Twingo, as well as the Megane and Scenic.

See, for example:
https://www.caroftheyear.org/

https://www.carwow.co.uk/renau...

Probably. Unfortunately, we have other companies with really crappy tech that refuse to die. My take is that IT is evolving far slower than it could because too many people make tons of money producing trash.

What makes you think they want to solve problems? My impression is that just want to tell everybody how great they are and they are trying to get that status by sabotaging everybody else.

The rich do not want more money. They want the poor to have _less_ money and especially less freedom, ideally no freedom at all. All of them have enough money that a bit more really does not matter. This is about oppression and, eventually, eradication or enslavement of large parts of the population so that they can feel superior. At least I do not see any other explanation that is left.

These are people with really bad personality defects and we unwisely gave them power. I mean, just look at how incapable and obviously insecure some of them are. The others just hide it better. These people know how defective they are. And they are now trying to rape the world to prove they are not.

That "efficient algorithm" may never materialize or may not materialize soon. Remember that AI has been an intense research topic for at least half a century. There are no easy to do things left, they have all been tried and most failed.

That said, we _know_ more computing power will not help. Lack of computing power is not what makes LLMs so bad. Those data-centers have no chance to ever be profitable. More likely, they will be a total write-off and maybe some not-so-pretty ruins.

I think they may even reach that at least for this decade. I mean, a known-defective and unfixable technology that costs far too much to run and maintain and has not made good on any of its promises and that has essentially (besides some cosmetics) stagnated? And then you pour in enough money to bring the economy down if anything goes wrong?

Sheer madness. Well, I hope this crash will defer the next AI hype (and there will be one as stupid people do not go away) by a few decades. Then this craze would at least have one positive outcome.

Funny thing is that these total cretins probably think they are doing a good job...

The proverbial elephant in the china-shop comes to mind. It is really surprising that Mozilla still exists with constant management mess-ups over the last 10 years or so.

Yep. This is not explainable below "complete incompetence" and "extreme arrogance" and, quite important for Japan, "extreme rudeness".

Probably just completely tone-deaf authoritarian asshole "management". I see no way this is not a drastic fail for their Japanese operation.