chaos_emergent

The reality is that none of that shit matters if you can build a product that people use and want to pay for. I would back someone who has made a dollar off of a product over someone who has built a great product that no one uses 100% of the time.

The reality is that you can make a successful business with okay engineering and great product insight. It's much more difficult to build a successful business with great engineering and poor product insight. Getting people to use and pay for what you've built gives you the product insight that you need.

> All thinking does is burn output tokens for accuracy

“All that phenomenon X does is make a tradeoff of Y for Z”

It sounds like you’re indignant about it being called thinking, that’s fine, but surely you can realize that the mechanism you’re criticizing actually works really well?

An alternative but similar formulation of that statement is that Anthropic has spent more training effort in getting the model to “feel good” rather than being correct on verifiable tasks. Which more or less tracks with my experience of using the model.

Most of the value I’ve gotten out of is has been observability. Graph and DAG workflow abstractions just help OTel structure your LLM logs in a clean hierarchy of spans. I could imagine figuring out a better solution to this than the whole graph abstraction.

Other than that I’m not too sure.

Have you considered that Claude set up a crontab that does that programmatically? Every 10 mins seems awfully, idk, regular.

The writer mentioned that people's intuitions about the distribution of land and where it's most valuable are wildly off, what exactly are people's intuitions that run counter to the data presented? It seems fairly intuitive to me that property values, as you get closer to an urban center

Human-driven research is also brute-force but with a more efficient search strategy. One can think of a parameter that represents research-search-space-navigation efficiency. RL-trained agents will inevitably optimize for that parameter. I agree with your statement insomuch as the value of that efficiency parameter is lower for agents than humans today.

It's really hard to imagine that they __won't__ exceed the human value for that efficiency parameter rather soon given that 1. there are plenty of scalar value functions that can represent research efficiency, of which a subset will result in robust training, and 2. that AI labs have a massive incentive to increase their research efficiency overall, along with billions of dollars and really good human researchers working on the problem.

I feel like the dropping cost of using AI doesn't tell the full story - I feel like I'm using agents easily a hundred to two hundred times more than I used chat interfaces. The build-out seems entirely reasonable if we believe that there's going to be a similar increase in usage through the rest of the economy as user interfaces are figured out for these models.

Probably not, they’re like four years old and they’re 2500 people at the company. My guess is that there are but a handful of PMs.

100% All of the people who are floored by AI capabilities right now are software engineers, and everyone who's extremely skeptical basically has any other office job. On investigating their primary AI interaction surface, it's Microsoft Co-Pilot, which has to be the absolute shittiest implementation of any AI system so far. As a progress-driven person, it's just super disappointing to see how few people are benefiting from the productive gains of these systems.

It’s almost like the iteration loop refines itself between checks notes in Sutton search and learning

Not at all, the limitation is software to get the model on the chip and executing correctly. My bet is that they had a FDE who specializes in the chip implement Spark’s architecture on device.

I think it's a beta so they're trying to figure out pricing by deploying it.

I grew up on a rural farm in California with a dial-up connection that significantly hampered my ability to participate in the internet as a teenager. I got Starlink installed at my parents' house about five years ago, and it's resulted in me being able to spend considerably more time at home.

Even with their cheapest home plan, we're getting like 100 Mbps down and maybe 20 to 50 up. So it's just not true at all that you would have connections that are a megabit or two per second.

Calling AI an unproven market is a wild statement. My mother and every employed person around me is using AI backed by Nvidia GPUs in some way or the other on a daily basis.

When you say “in” them, are you referring to their training data, or their model weights, or the infrastructure required to run them?

Isn't the struggle of sifting through a labyrinth of physical books and learning how and where to find the right answers part of the learning process?

I would argue a machine that short-circuits the process of getting stuck in obtuse books is actually harmful long term...

BuildBetter | Senior Engineer (Backend, Full-stack) | Full-Time | Remote (US) | 170-200k base + equity | Visa Sponsorship not provided

BuildBetter is the product operations platform for customer-led teams. We've built an AI-first product that lets our customers, like Rappi, n8n, Posthog, and Brex automate the operations work that lets them provide their customers with world-class product experiences.

We're looking for two senior engineers to join the team; our headcount has remained small while growing revenue, and we're now bottlenecked by development speed in a way that will make you invaluable. You'll have real ownership over your code from product ideation to customer support, and we'll be making substantive product updates weekly. We're looking for team players who are as fun to work with as they are productive, and are willing to compensate these two rolls commensurately - our aim is to be in the 95th percentile of equity compensation, and we expect that in exchange we'll have value-aligned employees who will help us grow the business.

Joining us will mean solving hard problems at the frontier of AI adoption, including:

- maintaining performance at scale

- building and maintaining rich frontend user experiences that maintain the standards of usability and delight that our customers know us for.

- building agent and workflow systems that automate real work for customers

- playing around with the latest models to figure out new experiences that we can offer our customers as cutting-edge models get released monthly

- building automation that's grounded in supporting product teams rather than automating them - our bet is that this is the likelier outcome out of the current generation of AI systems

Stack: Postgres / TS (Nest + React) / Hasura / EKS / S3 / LLM APIs / Python

Benefits:

- unlimited PTO

- $100/mo education stipend

- flexible hours: we measure productivity in terms of outcomes, not time, so as long as you're able to stay aligned, productive, and in regular communication, we don't really care what hours you work

- health insurance

Please reach out to: nikhil (at) buildbetter (dot) app

Most QA, most analyst positions, a good chunk of the kludge in intellectually challenging jobs, like medical diagnostics or software engineering, most administrative work, including in education and in healthcare, about 80% of customer success, about 80% of sales, are all within striking distance of automation with current-generation LLMs. And taht's entirely ignoring the 2nd-order effects in robotics and manufacturing.

Just a moment to reflect on how much freaking leverage computers give us today - a single permission change took down half the internet. Truly crazy times.

BuildBetter | Senior Engineer (Backend, Full-stack) | Full-Time | Remote (US) | 170-200k base + equity | Visa Sponsorship not provided

BuildBetter is the product operations platform for customer-led teams. We've built an AI-first product that lets our customers, like Rappi, n8n, and Posthog, automate the operations work that lets them provide their customers with world-class product experiences.

We're looking for two senior engineers to join the team; our headcount has remained small while growing revenue, and we're now bottlenecked by development speed in a way that will make you invaluable. You'll have real ownership over your code from product ideation to customer support, and we'll be making substantive product updates weekly. We're looking for team players who are as fun to work with as they are productive, and are willing to compensate these two rolls commensurately - our aim is to be in the 95th percentile of equity compensation, and we expect that in exchange we'll have value-aligned employees who will help us grow the business.

Joining us will mean solving hard problems at the frontier of AI adoption, including:

- maintaining performance at scale

- building and maintaining rich frontend user experiences that maintain the standards of usability and delight that our customers know us for.

- building agent and workflow systems that automate real work for customers

- playing around with the latest models to figure out new experiences that we can offer our customers as cutting-edge models get released monthly

- building automation that's grounded in supporting product teams rather than automating them - our bet is that this is the likelier outcome out of the current generation of AI systems

Stack: Postgres / TS (Nest + React) / Hasura / EKS / S3 / LLM APIs / Python

Benefits:

- unlimited PTO

- $100/mo education stipend

- flexible hours: we measure productivity in terms of outcomes, not time, so as long as you're able to stay aligned, productive, and in regular communication, we don't really care what hours you work

- health insurance

ALSO LOOKING FOR: Senior Product Designer | $105k - $180k USD | Design + code in HTML & CSS (Tailwind)

For engineering roles, please reach out at: nikhil (at) buildbetter (dot) app

For the designer role, please reach out at: adam (at) buildbetter (dot) app

Yes exactly, my theory is that the novelty of a new generation of LLMs’ performances tends to cause an inflation in peoples’ perceptions of the model, with a reversion to a better calibrated expectation over time. If the developer reported numerical evaluations that drifted over time, I’d be more convinced of model change.

Customer here, you're tripping. Recall provides transcription as an auxiliary service, not their core value prop.

Recall is, at its core, an API for bot recording. As someone building an application that relies heavily on conversational data, recording meetings is really important. Recall makes that process as easy as an API call, standardized across various meeting platforms. It's a huge PITA to set up infrastructure to get bots to join meetings that handle each platforms' proclivities, encoding and storing video data, etc.

The transcription service is just something they do to make transcribing recordings - one of the most common first post-processing steps for any conversational data - easier and lower friction.

I mean, Claude has had MCP use on the desktop client forever? This isn't a new problem.

knowledge cards at the top of Google results have been around for at least 12 years, I'd interpret the LLM-based responses as an iteration of a feature that's been around for a while rather than mimicking a competitor.

I think the problem is that methane is 20x more powerful a GHG than CO2

Author advocates for building general purpose systems that can accomplish goals within some causal boundary given relevant constraints, versus highly deterministic logical flows that are created from priors like intuition or user research.

Parallel with the bitter lesson being that general purpose algorithms that use search and learning leveraged by increasing computational capacity tend to beat out specialized methods that exploit intuition about how cognitive processes work.

Thank you, that was a wild thing to learn!

> This is likely because LLMs are typically trained for only a single epoch over massive datasets, which is in contrast to the multi-hundred-epoch training regimes for which dropout was first introduced.

Wait, is this true? That seems like a wild statement to make, relatively unsubstantiated?

Au contraire, ANNs are precisely the decomposition of larger problems into smaller ones.