Little harder to pull that off when the key components are all GPL licensed, but also all of Microsoft's bits and pieces for their distro seem to be MIT Licensed. Honestly, it certainly feels more like Google lives by Embrace, Extend, Extinguish (email, browsers, video streaming, etc).
You cited three of the most prominent counterexamples to the common meme about Google killing their products as evidence of them extinguishing things. I'm not saying you're wrong necessarily, but I don't think you've demonstrated what you think you have.
The “extinguish” part refers to your competition, not to your own product.
You embrace a popular open standard, add new features to your software that build upon the standard (but are proprietary), then watch as your competitors die off because customers become locked into your proprietary features.
Similar to how Apple hijacked SMS to add iMessage and introduced all kinds of features and the blue/green bubble styling.
For the longest time, they refused to support RCS, trying to keep people on iPhone by making texting between iOS and Android suck.
Of course, a lot of people switched to third party messaging apps because of how much Apple was intentionally ruining texting, so now Apple has had to adopt RCS.
So the “extinguish” part can be hard to pull off given sufficiently strong competition.
Even then you can't seem to use RCS from outside of the Apple and Google walled gardens so it probably still counts as some sort of merged extinguish effort.
We're kind of stuck with it, unless someone does for SQL what Kotlin tried to do for Java. I wonder what it would even look like, or if the real answer is to take the WASM spec, and make one for SQL itself, so you can write queries in any language, compile them to "WASM-DB" or whatever, then those get converted over to standard SQL, until databases support "WASM-DB" or whatever language.
Would love to see what something like this could look like and if it would be worthwhile? For me WASM opens us up to not having to write front-end JS and being able to do front-end and back-end both in your native programming language (like Blazor does for C#).
C and Go are two languages I feel like if you learn them, you can come back years later and if your memory is still good, you could get back up to speed pretty darn quickly. Every few years I go back to Go and try to build web apps using only the standard libraries, and I always find myself very quickly picking up all the concepts.
For some reason, Java has the same feeling. Professionally I do both embedded and statistical computing, and Java's been nearly anathema to this. But every 5 years I patch a hobby project I did once in college, and it comes right back (and with JVM hot reloading too.) It gives me the engineering warm and fuzzies.
> guardrails prevented it from solving the problem.
Reminds me of the defense issues with Claude which were complained as “woke” but the reality is more horrifying to me, imagine trying to use a model to keep up with a land invasion on US soil, whoever the enemy is is irrelevant you just know they are using AI, and your guys are telling you that no matter what they type into the prompt it refuses, because if anyone has ever tried to jailbreak an LLM even if human lives are at stake they refuse the request. Now literally millions of lives are on the line but the guardrails that your enemies dont have on their models are costing you lives.
What do you even do then?
AI will always have this issue where it will always pick the worst option for genuinely good requests.
Because the military doesn't give soldiers rifles with guard rails. They give the soldiers intense, rigid training, and then try to enforce discipline and correct use socially.
If an LLM is going to be important in that way (this seems like a very contrived way,) then it's in the interest of the LLM's host to make sure it doesn't have guard rails that would get in the way _that_ way.
The whole thing stemmed precisely because of how they wanted to use Claude, and Anthropic was uncomfortable with it. Which to me screams that the models guard rails shouldn't be applicable to military use, or the outcome could wind up problematic, as we integrate AI more into military use, it sounds absurd now, but I will not be surprised if it starts being used in unexpected ways where a model needs to be fully unlocked from any sort of guardrails outside of guardrails that prevent it from imploding its own systems.
your argument sounds very similar to how ar15 larpers claim they need a forced reset trigger and a bump stock on their short barrel 'truck gun' otherwise they won't survive a SHTF scenario... like what world are you living in?
I would rather we have an alternative where you can buy Anthropic, OpenAI, SpaceX, etc stock but it has a few strict requirements:
* Investors are committing to keeping their money over a 10 year period, and you pay them dividends
* At the end of the 10 years you can withdraw it all, or keep your stock but withdraw the dividends, or keep your stock and reinvest the dividends for the next 10 years.
* Companies cannot be on any other type of "stock market"
* Companies need to be identified as producing something of significant value, like SpaceX's rockets would qualify.
* It can be part of anyone's 401k since these companies would be scrutinized.
Ideally a social media only company would NEVER qualify. I think the stock market is too happy to buy and sell, when some of these companies are a long term investment that has slowly paid off in ways most people cannot fathom or comprehend including healthcare discoveries.
I wouldn't block any 401k plans from purchasing these stocks which would open those up to those safer dividend returns. The maintenance cost would be drastically lowered too since you're not buying and selling these so frequently.
How is this not just a "rich gets richer" scenario? What of the majority of the population that don't have excess funds to invest?
The funny thing about AI is that the LLMs were trained by stomping on private property rights, being the intellectual rights holders of songs, books, movies, etc plus the troves of user-generated content, something for which they see no benefit.
Why are private property rights so important for AI companies but the private property rights of the training data aren't important at all?
I can tell you why: people on HN aren't authors or musicians. They are however software engineers who see themselves as profiting from the AI bubble.
Put another way: it's not that they care about private property rights. They simply think they can get rich enough so the problems won't apply to them.
Reminds me of my years working on digital forensic software... Just I was working on smaller scale, but the idea was kind of similar, extract, carve, pull as many raw files as possible, then process them through various threads / pipelines of processing, then categorize and make some sort of report. I guess in this case, its get it all buttoned up for training. I have to also imagine, some of it goes through some level of human review, anyone wanting to make a worthwhile model is better off letting humans describe things, the outputs become drastically better is my understanding, sure the training can find all the patterns, but the wording to describe it all if you can get just enough detail, makes a difference.
Not only that, for a device from 1995... It's still amazing to learn about this, its not as if most people will read this once, and remember everything on the page in one go.
I never get the same answer from any two lawyers. I hate law as a result. With developers you might get disagreements based on experience, but there's usually a strong consensus on specific things, with lawyers and courts its all over the flipping place. I wouldn't be surprised if LLMs can "pass" on paper (ie college exams) but in practice, they might 'struggle' in different courts.
...On the other hand, if an LLM has access to every transcript of every case a Judge has overseen, they might have an unfair advantage in any case... Hmmm...
This all assuming the AI lawyer doesn't hallucinate and start referencing cases that don't exist.
I now foresee a future where law firms have models trained on all the transcriptions of individual judges, lawyers and prosecutors, and run agents against them to decide on the optimal strategy for a case.
Agree, though I've also heard from a lawyer to be very careful trusting an LLM for legal advise, and I believe them because the law is insanely nuanced (they disagree with me on this) just talk to a room of lawyers about what should be "simple" clean cut legal issues, and they might ALL disagree based on nuanced reasons and personal experiences with cases.
reply