This is what worries me. People become dependent on these GenAI products that are proprietary, not transparant, and need a subscription. People build on it like it is a solid foundation. But all of a sudden the owner just pulls the foundation from under your building.
But these products are all drop in replacements for each other. I've recently favored Codex more than CC, just because rate limits got mildly annoying. I really didn't have to change anything about my workflow in doing that.
> But these products are all drop in replacements for each other
For now. That doesn't really change the risk, that just means they are all hyper competitive right this moment, and so they are comparable. If one of them becomes king of the hill, nothing stops them from silently degrading or jacking prices.
The only shield is to not be dependent in the first place. That means keeping your skills sharp and being willing to pass on your knowledge to juniors, so they aren't dependent on these things.
Of course, many people are building their business on huge AI scaffolding. There's nothing they can do.
I'm curious - why for now? This stuff is practically commoditized. Trying to think of anything that ever successfully got back into proprietary land from there.
It doesn't look commoditized to me, it looks subsidized. It looks like everyone is trying to be "the one" and running as competitively as possible until the others fail. Commoditized would imply these services are all going to mellow into a stable state and mostly compete on price. I don't think that's happening. These aren't paper clips, they are courting governments and trying to pull the ladder up behind them. That's why both Anthropic and OpenAI are preaching doomsday and trying to build a moat with regulations.
Fair. I have high hope for local inference, feel like right now it is simply cost prohibitive to get the hardware. It will be interesting to see what happens.
The thing is that AI is still more akin to a glorified autocomplete than something that can really supersede your skills. Proprietary model suppliers are constantly trying to obscure this basic underlying fact, without much success (much of the unpredictable shifts you see in proprietary AI behavior ultimately boils down to this); so it becomes far more crystal-clear when using open models that really are a pure commodity.
yeah, I think there's the marketing and then there's the actual true utility. AI isn't a better computer program. It's not going to be able to do everything you want autonomously. But, it's pretty good at some stuff!
>Of course, many people are building their business on huge AI scaffolding.
It's similar in the way many businesses transitioned their scalability etc. to 'the cloud' starting a couple of decades ago.
It's a combination of loss of control and abdication of responsibility. They can claim to the customer the reason the service went down is now Microsofts or Amazons etc. etc. fault. Ultimately the end-user was the one that ended up losing.
It was a choice. There was something they could do - and keep everything in house, although cost-competiveness becomes an issue at some point and you get priced out of your target market. Everyone loses except for the cloud computing (or now AI) providers.
At least some of the investors in this tech are hoping for a monopoly position. They'd like to outspend the competition to get an insurmountable lead, at which point they can set their price.
But, so far, competition remains fierce. Anthropic still has the best tools for writing code. That lead is smaller than it's ever been, though. But, honestly, Opus 4.5 is when it got Good Enough. If Anthropic suddenly increased prices beyond what I'm willing to pay, any model that gives me Opus 4.5 or better performance is good enough for the vast majority of the work I do with agents. And, there are a bunch of models at that level, now maybe including some discount Chinese models. Certainly Gemini Pro 3.1 is on par with Opus 4.5. Current Codex is better than Opus 4.5 and close to Opus 4.7 (though I won't use OpenAI because I don't trust them to be the dominant player in AI).
I often switch agents/models on the same project because I like tinkering with self-hosted and I like to keep an eye on the most efficient way to work...which models wastes less of my time on silly stuff. Switching is literally nothing; I run `gemini` or `copilot` or `hermes` instead of `claude`. There's simply no deep dependency on a specific model or agent. They're all trying to find ways to make unique features for people to build a dependence on, of course, but the top models are all so fucking smart you can just tell them to do whatever thing it is that you need done. That feature could probably be a skill, whatever it is, and the model can probably write the skill. Or, even better, it could be actual software, also written by the model, rather than a set of instructions for the model to interpret based on the current random seed.
Currently, the only consistent moat is making the best model. Anthropic makes the best model and tools for coding, but that's a pretty shallow moat...I could live with several other models for coding. I'll gladly pay a premium for the best model and tools for coding, but I also won't be devastated if I suddenly don't have Claude Code tomorrow. Even open models I can host myself are getting very close to Good Enough.
It feels more and more like OpenAI/Anthoropic aren't the future but Qwen, Kimi, or Deepseek are. You can run them locally, but that isn't really the point, it is about democratization of service providers. You can run any of them on a dozen providers with different trade-offs/offerings OR locally.
They won't ever be SOTA due to money, but "last year's SOTA" when it costs 1/4 or less, may be good enough. More quantity, more flexibility, at lower edge quality. It can make sense. A 7% dumber agent TEAM Vs. a single objectively superior super-agent.
That's the most exciting thing going on in that space. New workflows opening up not due to intelligence improvements but cost improvements for "good enough" intelligence.
You can run local models on junker laptops for specific tasks that are about as good as last years SOTA. If the manufactured compute hardware shortage wasn't happening a lot more people would be running two months ago SOTA locally right now. Funny thoughts...
Open Source isn't even within 50% of what the SOTA models are. Benchmarks are toys, real world use is vastly different, and that's where they seriously lag.
Why should anyone waste time on poorer results? I'd rather pay my $200/mo because my time matters. I'm not a poor college student anymore, and I need more return on my time.
I'm not shitting on open weights here - I want open source to win. I just don't see how that's possible.
It's like Photoshop vs. Gimp. Not only is the Gimp UX awful, but it didn't even offer (maybe still doesn't?) full bit depth support. For a hacker with free time, that's fine. But if my primary job function is to transform graphics in exchange for money, I'm paying for the better tool. Gimp is entirely a no-go in a professional setting.
Or it's like Google Docs / Microsoft Office vs. LibreOffice. LibreOffice is still pretty trash compared to the big tools. It's not just that Google and Microsoft have more money, but their products are involved in larger scale feedback loops that refine the product much more quickly.
But with weights it's even worse than bad UX. These open weights models just aren't as smart. They're not getting RLHF'd on real world data. The developers of these open weights models can game benchmarks, but the actual intelligence for real world problems is lacking. And that's unfortunately the part that actually matters.
Again, to be clear: I hate this. I want open. I just don't see how it will ever be able to catch up to full-featured products.
Unless you are getting outside of your comfort zone and taking a month off from your $200 subscription, every other month, I can’t see how you can make the universal claim that the open weights models are all 50% as good. Just today, DeepSeek released a new model, so nobody knows how that will compare, a week ago it was Gemma 4, etc. I’m okay with you making a comparison, but state the model and the timeframe in which it was tested that you are basing your conclusions on.
I think that there will come a point when open source models are "good enough" for many tasks (they probably already are for some tasks; or at least, some small number of people seem happy with them), but, as you suggest, it will likely always (for the forseeable future at least) be the case that closed SOTA models are significantly ahead of open models, and any task which can still benefit from a smarter model (which will probably always remain some large subset of tasks) will be better done on a closed model.
The trick is going to be recognizing tasks which have some ceiling on what they need and which will therefore eventually be doable by open models, and those which can always be done better if you add a bit more intelligence.
> Benchmarks are toys, real world use is vastly different...Why should anyone waste time on poorer results? I'd rather pay my $200/mo because my time matters.
This kind of rhetoric is not helpful. If you want to make a point, then make one, but this adds nothing to the conversation. Maybe open source models don't work for you. They work very well for me.
> Open Source isn't even within 50% of what the SOTA models are
Who said so? GLM 5.1 is 90% Opus, at least. Some people quite happy with Kimi 2.6 too. I did not try Deepseek 4 yet but also hearing it is as good as Opus. You might be confusing open source models with local models. It is not easy to run a 1.6T model locally, but they are not 50% of SOTA models.
> Benchmarks are toys, real world use is vastly different, and that's where they seriously lag.
I'm not disagreeing per-se but if you think the benchmarks are flawed and "my real world usage" is more reflective of model capabilities, why not write some benchmarks of your own?
You stand to make a lot of money and gain a lot of clout in the industry if you've figured out a better way to measure model capability, maybe the frontier labs would hire you.
amazing how often is this repeated on here are some sort of a gospel SWEs pass down to one another to continue this charade. I have worked in this industry for 30+ years on countless projects, last decade+ as consultant - at every single project (every single one) programming time was the limiting factor. there is a whole industry inside our industry dealing with “processes” and “how to estimate” (apparently we are incapable of doing that) and whatnot, all because the actual programming time is always a limiting factor and there isn’t an even close 2nd
Agreed, it's very strange. I'm sure there are many projects that are like they describe, but it's certainly not all of them. I have worked as a game dev for over 20 years, and probably 75% of that time my team and I have been coding. AI has been an incredible game changer for me over the past 6 months or so (I was using it quite a bit before then, but the capability became much higher lately). I actually have some free time in my days now while still hitting milestone dates, instead of endless crunching.
What counts as programming time ? Writing ? Reviewing ? Compiling ? Debugging ? It also depends the industry. From idea to production, the limiting factor is not always writing the code, and in my experience (15years in fintech) it almost has never been. Discussion, alignment, compilation, heavy testing pipelines, shipping, all of this on a 30million line monorepo.
On a greenfield 10k line repo, yes, AI really shines. In other cases, it’s currently just a helper on very specific narrow tasks, that is not always programming.
No, its the number of debug cycles you need to solve said problems. That's the major attribute that controls dev time. And models require far more than I need. You are paying money to take longer and produce worse code. If its different for you, that's a you problem.
> Open Source isn't even within 50% of what the SOTA models are.
When was the last time you used any of them? Because, a lot of people are actively using them for 9-5 work today, I count myself in that group. That opinion feels outdated, like it was formed a year ago+ and held onto. Or based on highly quantized versions and or small non-Thinking models.
Do you really think Qwen3.6 for a specific example is "50%" as good as Opus4.7? Opus4.7 is clearly and objectively better, no debate on that, but the gap isn't anywhere near that wide. I'd call "20%" hyperbole, the true difference is difficult to exactly measure but sub-10% for their top-tier Thinking models is likely.
Their opinion is also behind on LibreOffice, too. I won't defend GIMP's monstrosity, but I finished a whole dissertation, do all my regular spreadsheet work (that isn't done via R), and have created plenty of visual mockups with LibreOffice. Plus, I don't have to deal with a spammy Windows environment.
Sure, we use Google Drive, too, but that's just for sharing documents across offices, not for everyday use. For that, the open source model is a clear winner in my book.
Qwen3.6 at which model size and quantization? I already think Opus 4.6 is usable but still dumb as bricks. A 20% cut off that feels like it would still be unusable. And that's not even getting to the annoyance of setting everything up to run locally & getting HW that can run it locally which basically looks like a Macbook M4 these days as the x86 side is ridiculously pricey to get decent performance out of models.
The largest qwen model is similar so I’m not sure what point you’re trying to make. The only ones available are the open weight ones which are the smaller variants and nowhere near within 20% of the closed frontier models.
The largest open models are within 20%; they're likely within 10%. Go actually try them and stop making outdated assumptions. You don't need to invest a lot of money either, just pick your favorite vendor, and send out a few prompts.
IMO It's a different and new model. We're engineers, and we're rich. It's not going to be good enough for us. But the much larger market by far is all the people who used to HAVE to work with engineers. They now have optionality; the pendulum is going to swing.
Also, this space will (and perhaps already is for some of us) be an arms race. Sure you can go local but hosted will always be able to offer more and if you want to be competitive, you'll need to be using the most capable.
If you pay someone $20,000 for labor, and they save 65 minutes worth of labor per day using a $200/mo Claude subscription, you are better off buying the Claude subscription.
Really? Are you using opus and letting it run for long periods? Curious as to what your workflow is.
The math is highly in favor of us using it at our company and we are paying API pricing. I don’t imagine there’s a lot of people using Claude without getting their money’s worth…?
Yes, recently I've been working on some research/ optimizatiom problem.
I would start claude in Yolo mode, tell it keep trying new ideas until it runs out of 1m context. (Every day I am giving it a hint to explore different directions as the sessions before)
Twice a day for a month, fits well into CC max plan.
I guess if I had to pay per token I would still use it but only for tasks where the value is clearer and immediate.
Everyone is arguing why I'm wrong or that I should have presented more data.
You've got the real insight with this claim.
This is the way the world is moving. Open source isn't even going where the ball is being tossed. There is no leadership here.
You're spot on.
If the cost to deliver a unit of business automation is:
A. $1M with human labor
B. $700k human labor + open source models
C. $500k human labor + $10,000 in claude code max (duration of project)
D. $250k with humans + $200k claude code "mythos ultra"
The one that will get picked is option "D".
Your poor college students and hobbyists will be on option "B". But this won't be as productive as evidenced by the human labor input costs.
Option "C" will begin to disappear as models/compute get more expensive and capable.
Option "A" will be nonviable. Humans just won't be able to keep up.
Open source strictly depends on models decreasing their capability gap. But I'm not seeing it.
Targeting home hardware is the biggest smell. It's showing that this is non-serious, hobby tinkery and has no real role in business.
For open source to work and not to turn into a toy, the models need to target data center deployment.
Yeah, I don't wanna shit on open source, there will certainly be uses for all different kinds of models.
The real money in this market, though, is going to be made in the C suite, and they don't really care about the model. They don't care if it's open source, closed source, or what it is. They don't want to buy a model. They're interested in buying a solution to their problems. They're not going to be afraid of a software price tag -- any number they spend on labor is far more.
Labor is something like 50%+ of the Fortune 500's operating expenses -- capturing any chunk of this is a ridiculous sum of money.
People pirate photoshop and office if they don't want to pay for it, making it as "free" as GIMP. If there is a free option people will use it. never underestimate the cheapskates.
I think the problem is that we're all waiting for the patented Silicon Value Rug Pull and ensuing enshittification, where there are a dozen tiers of products, you need 4 of them, and they now cost $2000/month. I want to hedge against that.
Maybe for folks who are deep into this, but it’s not exactly accessible. I tried reading up on it a couple of months ago, but parsing through what hardware I needed, the model and how to configure it (model size vs quantization), how I’d get access to the hardware (which for decent results in coding, new hardware runs $4k-$10k last I checked)—it had a non trivial barrier of entry. I was trying to do this over a long weekend and ran out of time. I’ll have to look into it again because having the local option would be great.
Edit: the replies to my comment are great examples of what I’m talking about when I say it’s hard to determine what hardware I’d need :).
For me the big hangup is the hardware. If I could find a simple guide to putting together a machine that I can run off an outlet in my home, I am sold. The problem is that I haven't found this yet (though I suppose I haven't looked very hard either).
$10K should be enough to pay for a 512GB RAM machine which in combination with partial SSD offload for the remaining memory requirements should be able to run SOTA models like DS4-Pro or Kimi 2.6 at workable speed. It depends whether MoE weights have enough locality over time that the SSD offload part is ultimately a minor factor.
(If you are willing to let the machine work mostly overnight/unattended, with only incidental and sporadic human intervention, you could even decrease that memory requirement a bit.)
As a typical example DeepSeek v4-pro has 59B active params at mostly FP4 size, so it needs to "find" around 30GB worth of params in RAM per inferred token. On a 512GB total RAM machine, most of those params will actually be cached in RAM (model size on disk is around 862GB), so assuming for the sake of argument that MoE expert selection is completely random and unpredictable, around 15GB in total have to be fetched from storage per token. If MoE selection is not completely random and there's enough locality, that figure actually improves quite a bit and inference becomes quite workable.
I've never seen reports of this kind of setup being able to deliver more than low single-digit tokens per second. That's certainly not usable interactively, and only of limited utility for "leave it to think overnight" tasks. Am I missing something?
Also, I don't know of a general solution to streaming models from disk. Is there an inference engine that has this built-in in a way that is generally applicable for any model? I know (I mean, I've seen people say it, I haven't tried it) you can use swap memory with CPU offloading in llama.cpp, and I can imagine that would probably work...but definitely slowly. I don't know if it automatically handles putting the most important routing layers on the GPU before offloading other stuff to system RAM/swap, though. I know system RAM would, over time, come to hold the hottest selection of layers most of the time as that's how swap works. Some people seem to be manually splitting up the layers and distributing them across GPU and system RAM.
Have you actually done this? On what hardware? With what inference engine?
I've been using local AI via LM Studio ever since I canceled my Claude subscription. It's obviously slower than Claude on my M1 Studio[†], but like someone else said, I use AI more like a copilot than an autopilot. I'm pretty enthused that I can give it a small task and let it churn through it for a few minutes, while I work on something alongside – all for free with no goddamned arbitrary limits.
[†] The latest Qwen 3.6 whatever has been a noticeable improvement, and I'm not even at the point where I tweak settings like sampling, temperature, etc. No idea what that stuff does, I just use the staff picks in LM Studio and customize the system prompts.
I love how it's just a tacit understanding that these companies' entire MO is to carve out a territory, get everyone hooked on the good stuff and then jack up the price when they're addicted and captured -- literally the business plan of crack dealers, and it's just business as usual in the tech industry.
I was recently introduced to the term "vcware", ala shareware or vaporware, to describe these products. "Don't use that, it's vcware, enshitification is coming soon."
Feasibility on commodity hardware would be the true watermark. Running high end computers is the only way to get decent results at the moment, but if we can run inference on CPUs, NPUs, and GPUs on everyday hardware, the moat should disappear.
You can already run inference on ordinary hardware but if you want workable throughput you're limited to small models, and these have very poor world-knowledge.
Indeed, I feel like we are in the early computer equivalent phase of AI, where giant expensive hardware is still required for frontier models. In 5 years I bet there will be fully open models we'll be able to run on a few $1000 of consumer hardware with equivalent performance to opus 4.7/4.6.
I think intelligence per compute will go up significantly in the coming years, while the cost per compute will drop significantly. No way to know for sure, so I guess we'll see
Sure, but local AI is still a black box. They can be influenced by training data selection, poisoning, hidden system prompts, etc. That recent Wordpress supply chain hack goes to show that the rug can still be pulled even if the software is FOSS.
Not really. The hardware requirements remain indefinitely out of reach.
Yes, it's possible to run tiny quantized models, but you're working with extremely small context windows and tons of hallucinations. It's fun to play with them, but they're not at all practical.
The memory requirements aren't that intense. You can run useful (not frontier) models on a $2-5K machine at reasonable speeds. The capabilities of Qwen3.6 27B or 35B-A3B are dramatically better than what was available even a few months ago.
Practical? Maybe not (unless you highly value privacy) because you can get better models and better performance with cheap API access or even cheaper subscriptions. As you said, this may indefinitely be the case.
This is why, despite enjoying all of this, I really want to focus on locally hosted models. If we don't host the technology ourselves, we're setting ourselves up for a hard fall down the line.
Until very recently, local models been little more than brittle toys in my experience, if you're trying to use them for coding.
But lately I've been running Pi (minimal coding agent harness) with Gemma4 and Qwen3.6 and I've been blown away by how capable and fast they are compared to other models of their size. (I'm using the biggest that can fit into 24gb, not the smaller ones.) In fact, I don't really need to reach for Claude and friends much of the time (for my use cases at least).
For the sake of argument if you build on AWS is that any more of a solid foundation? You're beholden to Amazon, unless you have the bandwidth to be able to DR immediately to another provider.
That is indeed a similar problem. Europe is now aware of this problem and starting to mitigate it, to reduce dependencies on big tech - how hard that may be.
Anthropic sells due to unrelenting pressure and unachievable demand > new owner cuts costs > models become worse > new owner sells > the capitalistic cycle wins > we, the people, suffer
You could say the same thing about your mobile phone bill. Most people still consider the benefits of roaming access to the internet greater than the downsides of being dependent on it.
As a 10y old, my father taught me about logical ports. I took a very large piece of paper and in a few days, I designed a tic tac toe "computer". It had LEDs that indicated the next computer move, based on the position of the pieces: every single possible state of the board led to a specific "next move" led. I do not think it actually would have worked, but of course I was very proud of my design at the time. Unfortunately, when I showed it to my teacher, he did not believe that I was serious. "This is a joke, right?" And that was it. Poor kid me... It did not discourage me however. I was a software engineer for a long time, and now I am a CS teacher. And I (try to) never ever discount the efforts of children.
That really hits home. I spent a couple weeks in primary school sketching my own blueprints for great inventions. Nothing that could've ever worked (I didn't know what a transistor actually was, but my machine certainly had a lot of them!), but in hindsight a good start for a curious tech-minded child - switches that opened/closed circuits, wires to connect the various imaginary lasers and electromagnets, and so on. On the back of the paper I scrawled documentation to remember what the darn thing was actually supposed to do (the biggest one? Save people who fall out of airplanes, which to my 9 year old mind was a big issue that needed to be solved)
One day my teacher noticed me doodling in the back, so she promptly grabbed all the "blueprints" I was so proud of, tore them up, and tossed them in the trash. I guess I get discouraged easier than you though, since I didn't design a thing for many years afterwards.
Thanks for sharing this. It is so sad! Sorry that there are people like that. The only thing we can do now, is be better people than those horrible teachers.
Oh god, what’s the deal with horrendous people becoming teachers? Lately, I’ve been, uh, “reminiscing” about how terrible adults were to kids when I was a kid (I’m gen X.)
It’s no wonder I turned my interest to the computer - it was only ever a jerk if I programmed it like that.
Same (GP). Schools were really unsafe places for children back then. It always strikes me of you see movies about schools in that period, that the story is often that children get horribly bullied and are called ugly, etc. I am glad my children grow up in better times.
Low barrier to entry and hard to get fired once you're in.
Rotten people put on a good face in the interview and then spread their misery around for decades to some of our most vulnerable. It happens in pretty much every unelected position in the public sector in my experience.
One of the things that got me in to "coding" when I was 9 years old was building tic tac toe in Excel, locking the window size to 3x3 cells and then implementing clicks as links to the next board state, with the "computer" having already played the next move. The whole sheet had every possible board state written out by hand.
Douglas Crockford nearly got cancelled because he qualified JavaScript as "promiscuous". People not knowing what the word means plus having a sense of urgency about sensitivity can be a dangerous combination.
> You're experiencing something real that the industry is aggressively pretending doesn't exist.
I agree with the article and recognize the fatigue, but I have never experienced that the industry is "aggressively pretending it does not exist". It feels like a straw man, but maybe you have examples of this happening.
His point is that the Orwellian way of surveillance is impossible to do in practice, and that a proper science fiction writer would have left the surveillance to machines. So I think his critique is about the art of SF writing, not about the prediction of surveillance itself.
Asimov missed the idea of the panopticon here, whereby control is self-enforced by the fear of being caught because you can be watched at any time, not all the time
Asimov's sci-fi has both hard and soft parts (especially his later works).
The main thing is that Asimov was more of a bright person(mensa member and professor) and good at making conjectures about development based on technology and it's impact on humans, rather than a great writer per-se (there's some famous interview from the 70s that makes a fair bit of things that weren't obvious at the time).
Like how he immediately goes to the feasibility of non-human total surveillance when concluding that the total surveillance of a population on the level of 1984 by humans is infeasible.
So this review is to large parts to be taken as an post-fact analysis about 1984 both from a standpoint of the predictions of it's conjectured future and an attempt to see _why_ conjectures failed (much of it, being attributed to Orwells need to expose his hatred for how infighting perverts socialistic causes).
> Asimov's sci-fi has both hard and soft parts (especially his later works).
Yeah I know Asimov. I actually really like his writings, which is why I am a bit surprised because this review is short-sighted and mean, and I think, misses the point.
> Like how he immediately goes to the feasibility of non-human total surveillance when concluding that the total surveillance of a population on the level of 1984 by humans is infeasible.
Right, but he still misses the point. As a physicist I can think about a dozen reasons why positronic brains make little sense. I accept this as some of the disbelief I have to suspend to get to the actual substance of the books. It’s no different. Me being a nerd does not mean that I have to be a jerk just because someone writes something I find implausible.
About feasibility, did Asimov even read the book properly? I remember quite well that telescreens were not permanently watched, but that wasn't necessary because the consequences of getting caught with "wrongthink" were terrible.
Near the end of the book Winston finds out that he was watched much more thoroughly than he thought. They read his private diary and carefully put the same mote of dust on top of the cover so that Winston wouldn't notice it had been opened.
The writeup does not mention Jeff Atwood (Stackoverflow founder) trying to convince Gruber to standardize markdown. Atwood approached him publicly in a series of blog posts, but Gruber kept silent, and if I remember correctly finally declined stating that he didn't want to spend time jumping through other persons' hoops. Although it sucks that markdown is not standardized, I still see this as an inspiring example of a person just doing what he wants to do.
It happened a bit differently; Atwood and friends simply came out with a standard document and called it "standard markdown", which Gruber then refused to endorse. Eventually after the series of blog posts and some back and forth they renamed the project "CommonMark", which it is still called today.
I am not sure (of course), but I think Atwood simply thought standardizing this format was so obviously valuable that he didn't consider Gruber might not want to work with him. In retrospect it's kind of nice that it didn't happen, it really keeps everyone incentivized to keep the format simple.
The linked post contains three cases of Markdown syntax (underscores) leaking into the text, where actual italics were likely intended. This is the most basic Markdown syntax element failing to work. The problem CommonMark is trying to solve is not adding new features (the only one they added to Gruber Markdown is fenced code blocks), but rather specifying how to interpret edge cases to ensure the same Markdown code produces the same HTML everywhere.
I understand the goal of the spec. In my experience once some spec document gets adapted widely enough, there's a strong incentive to add new features to it, which renderers would then be compelled to implement. Before you know it, MD is a complicated spec that doesn't serve its original purpose.
In this case a few minor edge cases is really not a big deal compared to that (in my opinion).
Another take on that is that Gruber is unable to sabotage a markdown standard from coming to exist, no matter how much of a tantrum he wants to have. I have no interest in listening to him about the topic, he's just in the way of the community and everyone is routing around the damage.
What Gruber has done is forced the spec to be called CommonMark, but as far as everyone except Gruber is concerned CommonMark is the markdown spec.
There are flavors that predate it like GFM, and extensions, but IMHO going forward it's CommonMark + possibly your extensions or it's not really Markdown.
