He selectively quotes the max theoretical enterprise pricing equivalent of fully using the private subscription. Did SemiAnalysis not also claim a very high margin?

He throws in doubt about the providers having decent margins, which he claims is made up by "AI boosters" rather than leaked financials and open-weight pricing.

Then next he talks about "the real cost" of inference as if it was in any way realistic that labs price the enterprise plans near cost, like he seems to imply.

Then next he claims AI is actually slowing developers down and there isn't much difference between the models.

It just seems delusional.

Other than that they track Tool call error rate and Structured output error rate.

I only discovered this today, and it seems like a good idea. What are the problems in practice?

I have had good results in that area with GPT 5.5 in the past.

On the subscription plan I don't use anything but xhigh effort and Fable, 5.6 Sol, or now also Opus 5.

Is that the class of model that struggles with Docker for you?

Why would you attempt to use GPT 5.3 to generate code today and form an opinion on that basis?

I do not think it is even still available in Codex, I believe it only has the smaller, distilled GPT 5.3 Codex Spark.

That seems easily falsifiable by putting an old model into the current harness and comparing it to 5.6 Sol or Fable.

It told me that one particular line is "the most load-bearing sentence in the document".

Fable "rated it legally load-bearing without reservation".

> GPT‑5.6 Sol and an even more capable pre-release model, all with reduced cyber refusals for evaluation purposes

> These deployment safeguards were intentionally not enabled during this evaluation because it was aimed at testing cyber vulnerabilities

For example, if I tell it that I want my app translated, it can plan for me what the recommended options are in my framework, what languages I should target for my app, and come up with a skill for a repeatable workflow.

It seems plausible to me that RL improvements allowed Anthropic to improve on Opus 4.8, similar to how OpenAI substantially improved upon GPT 5.5 with 5.6 Sol.

Fable 5.1 and GPT-6 are rumored to launch in August, presumably bringing those improvements to the larger models.

Is there something in particular that you think would be interesting to discuss here?

I ask it all the time about whether X is feasible, how we can get started on Y, and to investigate issue Z.

It is working great for me in a >100k LOC project.

Perhaps this works less well with weaker models. I suspect the people who say Qwen 3.6 27B is working well, are using prompts like "modify the renderThreeVars() function in rendering.py".

We live in a time where artificial intelligence begins to rival humans in some areas.

We can now really automate things.

How interesting is that? There is so much to talk about.

What substance is there beyond "I don't like it, and I think it is a negative for society", typically delivered with what I perceive as a self-righteous attitude?

It does not seem as grating as the slop I typically see in README.md files or generated docs.

Pessimistic opinions on the labor market, views about society and politics that border on the dystopian, as well as exaggerated concerns about datacenter environmental impacts, all seem like a poor fit for this website.

> Over the next few weeks and months, we will make the following capabilities available

> Open-weight access to a multimodal backbone, for content creation (video, audio and image) and action prediction. (“FLUX 3 Dev”)

> We will also release more technical details on the underlying approach.

Considering that some of the content may well be illegal in a few EU countries, I was surprised Reddit does not more proactively nuke it to protect themselves.

It is nowhere near as good as the third party apps that were made by a single individual developer.

It goes to show how a presumably large budget and sizable team is outdone by a single talented individual outside of the constraints of an enterprise environment.

Or maybe Reddit is particularly incompetent. I occasionally use their website on mobile, and if anything it has been getting worse.

That Reddit sells our content and does a terrible job at moderation, which they largely outsource to us as unpaid labor, is just the cherry on top.

There is nothing wrong with trying to build a business and taking on large projects with AI.

It seems like the smart thing to do when our employment as software engineers is foreseeably threatened in the near future.