This is more to encourage people to try out adding AI into their product, which is a totally different flow and experience from using AI to build the product.
10,735 karma · joined June 8, 2009
Previously senior developer at the MBTA, and before that co-founder and Lead Engineer at CoachUp, a Ruby on Rails start-up in Cambridge, MA (TechStars Boston F12).
Feel free to shoot me an email about anything: gabe@durazo.us.
The postings on this site are my own and do not represent PagerDuty's positions, policies, strategies, or opinions.
This is more to encourage people to try out adding AI into their product, which is a totally different flow and experience from using AI to build the product.
So I was wondering if the same applied here. My initial reaction is that "light" meaning "illuminated" and "light" meaning "not heavy" is that this is just one of those coincidences, since I can't see how that one word relates the two concepts. But given that Japanese also does the same, it makes me wonder if maybe there is some sort of relationship. Maybe it was a common belief historically that the sun was not very heavy or something like that. Or maybe not and it's just a rare double coincidence!
I don't get the "but what if it's somewhat - not infinitely - bad" argument? The known good is so dramatically good that we'd need a shocking amount of unknown bad to have made it not worth it. But if I'm choosing between mosquitoes, and say... fish, frogs, and a variety of birds, yes, get rid of the mosquitoes.
> Fun fact: “Akari” means light in Japanese, both as illumination and as lightness.
Wow! I love it when two languages have matching homonyms. I always wonder "what does it mean ?". Like, in English "second" to mean the unit of time and the place after "first" feels like two distinct usages. But it's the same in Spanish ("segundo"), which points to a deeper connection (and isn't so unsurprising given the shared roots of the languages): the unit of time is called that because it's the second subdivision of an hour.
The two usages of "light" in English are so distinct I'm quite surprised it's like this in the entirely unrelated Japanese language. Is this just a big coincidence, or does anyone know any deeper reasons here?
Can someone help me understand this? I might have an out of date mental model of how these things work.
Fundamentally, LLMs output tokens 1 at a time, generating the next token from all the previous. And as the context window gets larger, this gets harder / slower / more expensive. So I get the idea of a maximum context window.
But I don't understand the point or meaning of an output token limit. I thought it was more a measure of price capping (since output tokens are more expensive) that a user could configure. I guess a model will keep generating tokens until it hits a "stop", so does this mean it's tuned to more aggressively produce output tokens? How does that fit into agentic loops. Are output token limits based on how long until it goes back to the user? Or does each "turn" of tool call, thought, tool call, thought, etc, get its own limit?
I pay for Consumer Reports, since you actually need a paid subscription to access the data, which makes sense structurally. I don't really know if the reviews are any good, though I've been relatively happy with it. I also paid a subscription for an independent supplement testing lab, but I forget what it's called at the moment.
I was quite surprised when I learned this (when Claude Code edited a file despite being in plan mode). I had previously assumed "plan mode" was a harness level concern, and restricted what tools could be used. I didn't expect that it was simply an addendum to the prompt.
But I think it gave me some good insight into where the heads are at of Anthropic employees building this. Basically, leaning on pushing everything to the model. That's why alignment is so important: a "sufficiently advanced" model doesn't require any tooling infrastructure around it, and I suppose Claude Code devs are targeting that future. I had previously thought there was more to a harness, but with "auto" mode these days, it seems like there's no desire to build in that direction.
So I'm unclear what you're actually saying and wondering if you've missed that. Are you saying that at every reasoning level it says Opus 5 beats Astra? I just compared Opus 5 high to Astra high and it has Astra as generally better than Opus.
Is this because the context is being saturated? How did you set it up?
Was the prompt something like "Here's the state of the board, you're white, your move, what do you do?" and then starting fresh each time? Or did it include the whole history of moves and board states and previous thinking tokens and so on? No judgment, just trying to add this data point (thanks for sharing!) to my mental model and understanding.
I'd be curious how it would work if it started fresh each time. My guess is it would never make an illegal move, although it may not actually play all that well.
So then, if it's not actually a "fly-by-wire" mechanical system, why would you choose that as the UX? It cramps my calf constantly keeping it partially activated and the pedal half way depressed. More often than not I'll relieve my foot by doing short stints of cruise control, but that's tedious to manage.
My dream set up would be the accelerator pedal is, well, an accelerator, and the brake is a decelerator, and if you're not pressing either the car maintains the current velocity. It feels like _most_ of the time that you're driving you're just trying to make it go the same speed, and then you can press whichever pedal you need if you're wanting to stop or start again.
I don't suppose any cars work like this? ICE vehicles more or less in the moment do but "coasting" does decelerate a small amount so that over time you have to push the pedal again.
https://support.toasttab.com/en/article/Are-customers-charge...
But even so, I still prefer the credit card instrument, used as a "charge card". For those that don't know, a "charge card" is common with businesses and is expected to be paid off every month. But it helps with managing your cashflow - e.g. you can keep your money in a savings account all month and make one transaction at the end - and it keeps your actual money from being at risk of fraud.
Most Europeans seem confused with how Americans use credit cards for everything, but about half of us use them as charge cards, paying them off every month.
That is a benefit, so I can see paying a small percentage for it, but I don't think that benefit alone justifies the 3-5% of a transaction that credit cards charge now.
Where I am it's increasingly common to see credit card fees when checking out. I get it, because merchants are being charged 3-5% of their total revenue. I wish there was a low fee credit card network that merchants didn't charge a fee for, so I could continue the simplicity of digital payments but opt out of this crazy Visa Infinite rewards accounting boondoggle.
> Snark noted, though I have no earthly idea why you felt it necessary
I think it's useful if you are, in fact, talking out your ass. It's hard to tell given your confident phrasing, but saying something which gives away that you don't actually have meaningful experience in the subject is worth calling out. It sounds like your interlocutor has a better sense of the state of the music world, and that was a way for them to indicate it.
You can probably identify the age of every commenter in this thread based on the particular role that sits with them.
The most recent I've seen otherwise is this Federal Reserve study[0] from 2022. It finds that the marginal return on swipes is actually slightly negative because of how juicy rewards have gotten, and 80% of their profitability comes from interest (with most of the rest fees):
> we find that, on average, the credit function makes up approximately 80 percent of the credit card profitability, whereas the contribution of the transaction function is slightly negative, as rewards and other expenses on credit card transactions outpace banks' interchange revenues.5 In addition, fees—in particular late fees—comprise approximately 15 percent of credit card profitability.
[0] https://www.federalreserve.gov/econres/notes/feds-notes/cred...
On the one hand, relative to our income it's not so important, but on the other it feels bad leaving $3k on the table.