Denzel

I think you and I have different definitions of “one-shotting”. If the model has to be steered, I don’t consider that a one-shot.

And you clearly “broke” the model a few times based on your prompt log where the model was unable to solve the problem given with the spec.

Honestly, your experience in these repos matches my daily experience with these models almost exactly.

I want to see good/interesting work where the model is going off and doing its thing for multiple hours without supervision.

CC confidently iterated until it discovered the issue. CC confidently communicated exactly what the bug was, a detailed step-by-step deep dive into all the sections of the code that contributed to it. CC confidently suggested a fix that it then implemented. CC declared victory after 10 minutes!

The bug was still there.

I’m willing to admit I might be “holding it wrong”. I’ve had some successes and failures.

It’s all very impressive, but I still have yet to see how people are consistently getting CC to work for hours on end to produce good work. That still feels far fetched to me.

It just felt really difficult to do both the engineering work while trying to do customer development at the same time.

The fact that OP has been able to do this for so long, while supporting a family, piqued my interest.

Also, you’re missing material capex and opex costs from a DC perspective. Certain inputs exhibit diseconomies of scale when your demand outstrips market capacity. You do notice electricity cost is rising and companies are chomping at the bit to build out more power plants, right?

Again, I ran the numbers for simplicity’s sake to show it’s not clear cut that these models are profitable. “I can sort of see how you can get this to work” agrees with exactly what I said: it’s unclear, certainly not a slam dunk.

Especially when you factor in all the other real-world costs.

We’ll find out soon enough.

> I'm not asking for magic, I'm asking where went the reliability we already had, at the prices we're already paying.

My god thank you! My partner and I have been talking about this for the past 2 years in the context of food service and delivery service industry.

Greater than 50% of all our restaurant orders are straight up wrong or missing items, whether it’s from local places, chains, or fast food restaurants.

The unreliability is staggering, especially because we’re paying so much more!

It’s gotten so bad that we’re done with certain services and establishments for good now, or we make sure to QC before leaving the restaurant to ensure everything is in the bag.

Even more ironic, this happened a couple weeks ago at Texas Roadhouse — the same restaurant I worked in decades ago as a teenager, so I remember the process we had to go through for to-go orders.

First, we’d take the order over the phone. We’d repeat the order back to the customer to confirm everything (1st QC). When the food came up in the window, we’d pack the food in bags, crossing off every item on the receipt before stapling it to the bag (2nd QC). When the customer came to pick up their food, we’d have to take every box out of the bag, show the customer the food, and confirm that everything they expected in their order was there (3rd QC).

No customer. Every left. With an incorrect order. Simple.

That process is gone now. We paid more and came home missing my partner’s meal. Wtf.

For simplicity’s sake we’ll assume DeepSeek 671B on 2 RTX 5090 running at 2 kW full utilization.

In 3 years you’ve paid $30k total: $20k for system + $10k in electric @ $0.20/kWh

The model generates 500M-1B tokens total over 3 years @ 5-10 tokens/sec. Understand that’s total throughput for reasoning and output tokens.

You’re paying $30-$60/Mtok - more than both Opus 4.5 and GPT-5.2, for less performance and less features.

And like the other commenters point out, this doesn’t even factor in the extra DC costs when scaling it up for consumers, nor the costs to train the model.

Of course, you can play around with parameters of the cost model, but this serves to illustrate it’s not so clear cut whether the current AI service providers are profitable or not.

As one of the WAU (really DAU) you’re talking about, I want to call out a couple things: 1) the LOC metrics are flawed, and anyone using the agents knows this - eg, ask CC to rewrite the 1 commit you wrote into 5 different commits, now you have 5 100% AI-written commits; 2) total speed up across the entire dev lifecycle is far below 10x, most likely below 2x, but I don’t see any evidence of anyone measuring the counterfactuals to prove speed up anyways, so there’s no clear data; 3) look at token spend for power users, you might be surprised by how many SWE-years they’re spending.

Overall it’s unclear whether LLM-assisted coding is ROI-positive.

- How much has your TTM reduce by? How did you measure?

- What's the net difference when you factor in token spend expenses?

- By how much can Anthropic increase prices before crossing over your break-even point?

You want your users to be like weight lifters. No lifter comes out the gym saying, “Man that was the best workout, felt so easy,” to the contrary, lifters use progressive overload to induce difficulty because that difficulty connects to the results they want.

For your users, you need some way to measure the outcome, so that you can show them, “hey look, that mild discomfort lead to more progress on what you care about,” and then you need to consistently message that some difficulty is good.

Mindset change takes consistency and time. Won’t happen over night. You’ll know you succeeded when students become aware of “hey, I’m not learning as well if it doesn’t feel difficult”, and then react by increasing the challenge.

In 2014, Facebook published a paper showing how they can manipulate users’ emotions with their news feed algorithm.

Facebook ran this test on 700k users without consent.

I deactivated my account the day I read that paper and never looked back.

DDD suggests continuous two-way integration between domain experts <-> engineers, to create a model that makes sense for both groups. Terminology enters the language from both groups so that everyone can speak to each other with more precision, leading to the benefits you stated.

When OP searches for Midjourney as a Midjourney user, Google’s algorithm infers he might want to consider an alternative because why would an existing user search for the product they’re already using.

We see evidence supporting this given no Midjourney ad showed up for a direct keyword match query; and only alternatives triggered.

This is kinda like Amazon retargeting you with alternative toasters after you just bought a new toaster. Most people think this is stupid. Well, the most likely cohort to buy a new toaster is a person that just bought one because they’re not satisfied with their purchase.

1) Help me understand what you mean by “pure compute providers” here. Who are the pure compute providers and what are their financials including pricing?

2) I already responded to this - platform power is one compelling value gained from paid API market share.

3) If the frontier lab you’re talking about is DeepSeek, I’ve already responded to this as well, and you didn’t even concede the point that the 80% margin you cited is inaccurate given that it’s based on a “theoretical income”.

It's worthwhile to note that https://github.com/deepseek-ai/open-infra-index/blob/main/20... shows cost vs. theoretical income. They don't show 80% gross margins and there's probably a reason they don't share their actual gross margin.

OpenAI is the easiest counterexample that proves inference is subsidized right now. They've taken $50B in investment; surpassed 400M WAUs (https://www.reuters.com/technology/artificial-intelligence/o...); lost $5B on $4B in revenue for 2024 (https://finance.yahoo.com/news/openai-thinks-revenue-more-tr...); and project they won't be cash-flow positive until 2029.

Prices would be significantly higher if OpenAI was priced for unit profitability right now.

As for the mega-conglomerates (Google, Meta, Microsoft), GenAI is a loss leader to build platform power. GenAI doesn't need to be unit profitable, it just needs to attract and retain people on their platform, ie you need a Google Cloud account to use Gemini API.

I appreciate all the experience and advice you’re offering on this thread! Take my feedback as a nitpick: as I was reading through your top post, my initial thought was “this isn’t true all the time” because I spent 6 years in 2 separate startups with significant and successful outbound sales where our ADV > $100k.

One company stayed private and profitable while driving revenue north of $80M/yr; and the other company sold enough long-term enterprise contracts to be acquired by a bigger $B company.

Context is king.

Not to mention there's literally people creating tech out here _today_ that's recreating _exactly_ what some Black Mirror episodes were talking about years ago. Like interactive chatbots model after dead people from voice samples, videos, and messages.

Grounding these conversations in an actual reality affords more context for people to evaluate your claims. Otherwise it’s just “trust me bro”.

And I say this as a Senior SWE who’s successfully worked with ChatGPT to code up some prototype stuff, but haven’t been able to dedicate 100+ hours to work through all the minutia of learning how to drive daily with it.

Within the context of the original discussion around whether self-driving is here, today, or not, I think we can definitively see it’s not here.

It’s safe to assume that a company’s ownership takes the decisions that they believe will maximize the value of their company. Therefore, we can look at Alphabet’s capital allocation decisions, with respect to Waymo, to see what they think about Waymo’s opportunity.

In the past five years, Alphabet has spent >$100B to buyback their stock; retained ~100B in cash. In 2024, they issued their first dividend to investors and authorized up to $70B more in stock buybacks.

Over that same time period they’ve invested <$5B in Waymo, and committed to investing $5B more over the next few years (no timeline was given).

This tells us that Alphabet believes their money is better spent buying back their stock, paying back their investors, or sitting in the bank, when compared to investing more in Waymo.

Either they believe Waymo’s opportunity is too small (unlikely) to warrant further investment, or when adjusted for the remaining risk/uncertainty (research, technology, product, market, execution, etc) they feel the venture needs to be de-risked further before investing more.

The promise has been that self-driving would replace driving in general because it’d be safer, more economical, etc. The promise has been that you’d be able to send your autonomous car from city to city without a driver present, possibly to pick up your child from school, and bring them back home.

In that sense, yes, Waymo is nonexistent. As the article author points out, lifetime miles for “self-driving” vehicles (70M) accounts for less than 1% of daily driving miles in the US (9B).

Even if we suspend that perspective, and look at the ride-hailing market, in 2018 Uber/Lyft accounted for ~1-2% of miles driven in the top 10 US metros. [1] So, Waymo is a tiny part of a tiny market in a single nation in the world.

Self-driving isn’t “here” in any meaningful sense and it won’t be in the near-term. If it were, we’d see Alphabet pouring much more of its war chest into Waymo to capture what stands to be a multi-trillion dollar market. But they’re not, so clearly they see the same risks that Brooks is highlighting.

[1]: https://drive.google.com/file/d/1FIUskVkj9lsAnWJQ6kLhAhNoVLj...