ndm000

I have been struggling with the same thing. Reading the other comments make me agree with you even more. The best I can come up with is two fold. First that humans will have jobs verifying the verifiers. That is, for important initiatives where something must be right, agents can be doing most if not all of the work, but it will still be a human’s job to investigate what was done and sign off on it. The second is that humans will still have to make agreements with other humans for resources, capital, and the like. Somethings will be scarce, others will become scarce. And humans will always own scarce things that other humans will need to use - and they won’t be making deals with machines.

Another frame I’ve heard is that people will only trust humans when it comes to their kids, health, and money. That’s the simpler view.

There is a comment in the intro of that book about the pinnacle of human labor being the simulation of consciousness. Very prescient for being written in the 60s.

Completely agree. All of the pieces are there and it's just waiting to be acted upon. I haven't seen any of the major players really doubling down on this, but would be so compelling.

I’ve felt the same way. It’s so inefficient to have two patterns - OLAP and OLTP - both using SQL interfaces but requiring syncing between systems. There are some physical limits at play though. OLAP will always take less processing and disk usage if the data it needs is all right next to each other (columnar storage) where as OLTP’s need for fast writes usually means row based storage is more efficient. I think the solution would be one system that stores data consistently both ways and knows when to use which method for a given query.

Have there been any updates to Claude 3.5 Sonnet pricing? I can't find that anywhere even though Claude 3.7 Sonnet is now at the same price point. I could use 3.5 for a lot more if it's cheaper.

One thing I have not seen commented on is that ARC-AGI is a visual benchmark but LLMs are primarily text. For instance when I see one of the ARC-AGI puzzles, I have a visual representation in my brain and apply some sort of visual reasoning solve it. I can "see" in my mind's eye the solution to the puzzle. If I didn't have that capability, I don't think I could reason through words how to go about solving it - it would certainly be much more difficult.

I hypothesize that something similar is going on here. OpenAI has not published (or I have not seen) the number of reasoning tokens it took to solve these - we do know that each tasks was thoussands of dollars. If "a picture is worth a thousand words", could we make AI systems that can reason visually with much better performance?

I think this is a key argument in how powerful AI can become. We may be able to create incredibly intelligent systems, but at the end of the day you can’t send a computer to jail. That inherently limits the power that will be given over to AI. If an AI accidentally kills a person, the worst that could be done to it is that it is turned off, whereas the owners of the AI would be held liable.

When steam and coal engines gave way to gas and electric engines in factories, it took decades before factories were reconfigured to adjust to the smaller sized engines that didn't require one major axle running through the entire factory. As a consequence the first gas engines were huge - over time they shrunk. I bet the same will happen with robotics, where humanoid will be the primary form factor at first for general tasks, then more efficient forms will emerge as processes are updated.

I wouldn't be surprised if the slowest part of the system is the API call to a legacy warehouse management system that takes several seconds to respond to get the next bin to target.

I've felt this way and have started celebrating inputs - the work involved in building something - rather than outputs (the sale, the deployment, etc.). Once it closes or the deployment is successful it's always onto the next thing immediately and there's no time to sit back and reflect on the hard work and enjoy the time spent with others in the process.

I think the compelling difference is truthfulness. There are certain people / organizations that I trust their synthesis of information. For LLMs, I can either use what they give me in low impact situations or I have to filter the output with what I know as true or can test.

After a recent 3 week driving trip through Europe, I can anecdotally back this up. It’s not even the manual transmission, but also the much smaller roads with no shoulder where at times you meet a car, have to slam on the brakes, and decide in the moment whether you or the other car will back up to a turn out. You have to pay constant attention and there’s little room for looking at phones, eating, etc. In the US we have such large roads and shoulders that you can zone out and are more easily tempted to take your eyes off the road.

Agreed. The same data privacy argument was used by people not wanting their data in the cloud. When an LLM provider is trusted with a company’s data, the argument will no longer be valid.

I think this is where it's going. Particular Mastodon servers will provide certain guarantees to their users and about their users. It's easy to envision a white-glove server + service that provides user screening, phone support, etc. at a price. Over time when certain celebrities join it becomes the popular server and sought after as a username destination.

I agree that taxation is part of money’s value, and there are also other factors. If money is backed by a commodity (i.e. gold) then part of its value comes from the value of the commodity.

For fiat currencies where new money is mostly created as debt, the common refrain is that its value is based on common trust. I think it’s more than that. Debt is an obligation between two entities regarding future outcomes - namely, the repayment of the loan plus interest. When looked at this way the value of money comes from humanities ability to affect future outcomes. If I take out a mortgage, the bank trusts me to repay based on evidence I’ve demonstrated, and I trust the bank due to the bank’s reputation, the laws that govern banking, etc.

I am not an economist and wouldn’t be surprised if these concepts are already in some theory I’m not aware of. I do think that the sentiment that money created out of nothing is not really correct - rather it’s money created out of expected future value.

I know nothing about optics. What is the effect that causes the 6 or 8 points of light of come off of bright objects? Does it have to do with the hex-shaped mirrors on JWT?

I’d like to understand more about the perception that the Economist is a liberal publication. I’ve heard this before and it’s hard for me to reconcile this view with the articles I read. In my opinion, the economist more often than not errs on the side of being comprehensive and including facts that may or may not be relevant. Some of those facts may point to liberal conclusions, sometimes not.

I also know that they’ve taken a strong conservative stance on gender-specific sports. If they were “unabashedly biased” that would not fly.

A couple of months ago I stopped checking the news except for once a week, and found weekly news summaries from the Economist to be extremely helpful. I found that I was more relaxed during the week, more contemplative about the events of the world, and had stronger reactions to things that occurred because I didn’t feel bombarded by so much information.

I am teaching my 10 year old Python using Nano, and he’s picked it up surprisingly quickly. When I wanted him to upgrade to VS Code, he preferred the terminal! Nano’s a great product, thanks for the hard work on it all.

At first I thought “all the female words must be in some other language”.

they bolt it onto Postgres

I am working with a company that is redesigning an enterprise transactional system, currently backed by an Oracle database with 3000 tables. It’s B2B so loads are predictable and are expected to grow no more than 10% per year.

They want to use DynamoDB as their primary data store, with Postgres for edge cases it seems to me the opposite would be more beneficial.

At what point does DynamoDB become a better choice than Postgres? I know that at certain scales Postgres breaks down, but what are those thresholds?

Reading through these posts, I had exactly the same thought. Why not have a language that takes the problems with SQL and abstracts them to something that is more easily organized into functions, modules, and packages, corrects some of the semantic problems, integrates well with source control, testing, deployment pipelines, etc. DBT is solving this problem well for data warehouses, but I do not know if a similar tool application databases.

Side question - do you need to reach a certain level of membership with HN to see the downvote button? I may be missing something, but seems that's only allowed for certain individuals.

I have worked with several companies that have their infrastucture on AWS but consider BigQuery or Snowflake for the serverless model they provide. This brings RedShift much closer to those options. I envision Redshift Serverless becoming the default options for most enterprises, mainly because it stays in the AWS ecosystem and you don't have to work with a different vendor and create different cost governance processes.

I beleive the real advantage AWS has here is in cost. Snowflake has positioned itself as price competitive with Redshift but this is primarily due to Snowflake's ability to scale on-demand, whereas prior Redshift versions required you to size for peak usage (RA3 helped with this). In my experience Snowflake is an order of magnitude more expensive if you compare similiar workloads and do not account for idle time. We will need to see the performance of a "Redshift Processing Unit" to be sure of the advantage, but even so AWS will be able provide significant downward cost pressure through this offering.

I’ve also found that when not spatchcocking, I have to make sure to tie the legs together. Doing so keeps the juices inside and keeps it really moist.

Completely agree, and it makes me wonder if we are at the edge of standardization for many types of applications. The desktop, tablet, and mobile form factors and use cases are well understood. Best practices amongst these major platforms are in place. The cloud is maturing (litmus test for this is the reinvent seems to be less exciting each year). New form factors, like AR/VR, don’t have the market share for broad changes to development practices yet.

I've seen this trend as well. There's this interesting double trend of upward costs on senior resources and downward costs on junior resources.

The North American equivalent is Latin America. I’m in consulting, and every new project has some set of near shore resources billed much lower than their North American counterparts. In most cases the only discernible difference is proficiency in English.

Is it normal to render using a CPU? Shouldn’t this test be done against GPUs instead?

The biggest problems I've had with Oracle databases are getting CDC data out of it. There's conflicting technologies with their own tradeoffs - LogMiner vs. Flashback, neither of which cover all use cases well, licensing is ambiguous, and docs are conflated with information you don't need.