data-ottawa

5 ms·

61.
by data-ottawa 4mo ago
> For years I could ask Google what song was playing and it would identify it. It was one of the most useful features it had. You'd hear a song in a store, on television or drifting over from a neighbour's backyard and Google w
62.
by data-ottawa 4mo ago
I feel this is a sort of inverse inspection paradox (the paradox that if you sample waiting time in a process, you’re more likely to sample a larger value). The LLM providers fine tune the models with some kind of information retrieval task
63.
by data-ottawa 4mo ago
Well I am definitely not using the models that I'm not able to access. So now the question is whether the capabilities of other models are worth their far cheaper token prices. Plus, are we at all confident Opus or GPT 5.5 aren't
64.
by data-ottawa 4mo ago
As a non-US citizen I guess this is the last money I pay to US companies for AI then. I can't help but wonder if it's now obvious that frontier AI work should not happen in the US. I can understand the KYC aspect of this, but at t
65.
by data-ottawa 4mo ago
I don’t know. I can write a simple query before Claude finishes reading, querying the semantic layer, checking my files, then writes a query that I have to approve, reads the results, hides them (ctrl+o usually works), and gives me a summar
66.
by data-ottawa 4mo ago
The cynic in me has learned one is measurable and can go on a slide deck, the other is vague and hard to measure.
67.
by data-ottawa 4mo ago
I think that’s mostly okay, I just worry about the times. It was like 2-3 seconds to search and pull context then generate the answer. I’m writing AI apps these days, and even pulling Gemini 3.5 flash on Google Cloud takes longer to get a m
68.
by data-ottawa 4mo ago
If you watched the keynote they added transparency controls, adjusted the refraction, removed the different window corner radii into one, and cleaned up sidebars and toolbars. They also updated icons to have more depth to them so you get a
69.
by data-ottawa 4mo ago
You could see jitter in the reflection on the MacBook as the guy typed, so that all looked great. The responses came in very fast though, so I’m sceptical that the latency is representative (or that they didn’t cherry pick results, but they
70.
by data-ottawa 4mo ago
Operator overloading has never been an issue for me, but terminating a line with a comma creating a tuple, or white space (including new lines) between strings to concatenate have cost me days of work over the years. I understand why those
71.
by data-ottawa 4mo ago
Control and performance, the same reasons why people write in low level languages today/yesterday instead of everything in Python or JS. Not all vibe coding is top down "Claude build X", it can be very specific implementation
72.
by data-ottawa 4mo ago
I had to do a graft and it wasn’t especially painful. I got painkillers but only used them on the first day. The recovery period sucked, I needed antibiotics and couldn’t eat anything solid for two weeks. If you can behaviourally prevent ne
73.
by data-ottawa 4mo ago
I plotted the sizes of shoes people buy from e-commerce data and there’s a similar bump at US 12 size, with a dip on 10 and 11, no such effect on women’s sizes.
74.
by data-ottawa 4mo ago
Does this maintain your browser extensions (and their settings)?
75.
by data-ottawa 4mo ago
I’ve never asked Claude to re explain things so many times with this. The language it chooses is bizarre, not quite technical enough to be precise, and too hand wavy to be useful. There also seems to be other issues with the sub agents and
76.
by data-ottawa 5mo ago
The nightmare was supposed to be “over in 3 years” 9 years ago. The foreseeable future is MAGA candidates with a coin flip odds of winning indefinitely.
77.
by data-ottawa 5mo ago
The AI product rollouts in the last two years have been some of the most aggressive and user hostile product rollouts in my entire life. All conventions and user centricity go out the window with AI feature launches lately. If you look at e
78.
by data-ottawa 5mo ago
I think my favourite part of the study is that the control group is just the class wait list. Unfortunately this study doesn’t control for luck.
79.
by data-ottawa 5mo ago
I mostly interact with their AI through bangs. An ending question mark enables fast answers, like Google’s AI summary. !ki sends your query to the assistant on light research mode. It runs a few searches against their index and summarizes t
80.
by data-ottawa 5mo ago
This is a philosophical question that goes back millennia. It just comes down to what sparks joy for you, and how much do you value that. I have an Eames lounger. It was absurdly expensive and doesn’t even have a recline lever. But, it spar
81.
by data-ottawa 5mo ago
When we had a book where only the homework problems changed in the new version we would pool together to buy one new copy and that person emailed out the homework questions. The rest of us bought used books at the start of semester used boo
82.
by data-ottawa 5mo ago
Yeah, I’m not against the idea, but I don’t have a Spotify membership. They’re simply not my music vendor of choice. Why should that lock me out of a reasonable reservation system? We have a massive centralization problem in these industrie
83.
by data-ottawa 5mo ago
I just read a pre-LLM thriller written almost exactly like this. This is dialed up to eleven, but this is a fairly common writing style. It switches my brain into skimming mode very quickly, as it reads quite padded.
84.
by data-ottawa 5mo ago
What thinking level were you using? In my testing, the minimal thinking mode hallucinated 2/3 times, which is pretty scary. The other modes weren’t as bad. I don’t have comprehensive data though.
85.
by data-ottawa 5mo ago
I'm seeing this too. I have a SQL agent and my tests with 3.5 are resulting in hitting query budget limits that have never been hit before. On average, to answer the same question, 3.5 is spending 10x more on SQL queries vs gemini-3-fl
86.
by data-ottawa 5mo ago
Because there's been a string of bad actors including OpenAI with incredibly inefficient scrapers. Previously captcha was just for spam limiting, but I actually looked at our system logs and about half of traffic was bad behaving scrap
87.
by data-ottawa 5mo ago
Anyone using this yet? I’m finding it very bad at instruction following vs 3.1. It calls tools it is told shouldn’t, and it loves calling tools. There’s a pretty strong bias towards its training vs system prompt instructions. Google’s relea
88.
by data-ottawa 5mo ago
I don’t know. Maybe today, but tomorrow? If you can sample points inside a volume, in theory you could do that with splat geometry. If someone figures out a way to pass in animation time to a sampler, sample along geometry/wireframe or
89.
by data-ottawa 5mo ago
reCaptcha is a pretty strong wall to allow only Google to index websites, especially now that you need device verification. Throw in Cloudflare too. There’s not much room to squeeze in when your competitors hold the keys to 15 million top w
90.
by data-ottawa 5mo ago
This all feels a bit hyperbolic. It doesn’t really say in the article search is going away. A lot of Google search is in the format of “company X”, then clicking the third link down (after two paid ads) to open company X’s website. (I have

More ›