blixt

16 ms·

181.
Show HN: A Chrome extension to quickly ask ChatGPT about the page you're on(github.com)
12 points by blixt 2y ago|4 comments
182.
by blixt 2y ago
I don't see any details on how API access to these features will work. This is the first true multimodal network from OpenAI, where you can send an image in and retain the visual properties of the image in the output from the network (
183.
by blixt 2y ago
GPT-4o being a truly multimodal model is exciting, does open the door to more interesting products. I was curious about the new tokenizer which uses much fewer tokens for non-English, but also 1.1x fewer tokens for English, so I'm wond
184.
by blixt 2y ago
I've been playing with the 8x22B version for several coding tasks today, and compared to DBRX and Command-R+, two other large and capable local models recently released, this WizardLM-2 is just a lot better. It follows the instructions
185.
by blixt 3y ago
I wonder how the reactions to this headset will play out. It's obviously going to be relatively clunky compared to expectations given that it's the first version. The marketing material doesn't really show the hassle of using
186.
by blixt 3y ago
Because before there was && the & was used to mean the same as && does today. This happened when the context was Boolean such as an if.
187.
by blixt 3y ago
It’s very smart, though using bounding boxes will most likely limit it to 2D contexts (and some head-on 3D contexts) since the text won’t follow the bounding box when perspective is involved. I’m sure it can be improved to support bounds th
188.
by blixt 3y ago
Since the example leads with: > Many features that make the Bosque IR amenable for automated reasoning involve simplifying and removing sources of irregularity in the semantics. I do find it strange that it both shows a bug in the logic
189.
by blixt 3y ago
Will await using actually perform the await at release, i.e. at the end of the scope? Could add an unexpected event loop tick around a curly brace (so if you schedule something after the await but before the end of the scope, it will look l
190.
by blixt 5y ago
Been hoping for more of this for years. We stare at diffs all day yet we have to accommodate the computer by understanding that the parenthesis it claims was changed wasn’t actually changed, there was just another set of parentheses added.
191.
by blixt 5y ago
The cache segregation is a bit of a nuisance when building sites that uses iframes on different domains to sandbox user content. For example, Framer (where I work) sandboxes each user project on a unique subdomain of framercanvas.com, which
192.
by blixt 5y ago
Also check out the Rabin rolling hash algorithm for smart chunking of files which can be useful to either dedupe dynamic content or efficiently sync files.
193.
by blixt 6y ago
I actually did this and while it follows the conversation quite well I think it’s really hard to make the argument that there was a decision early in the text what animal it thought of. That would only be possible if there were hidden embed
194.
by blixt 6y ago
I tried running the riddle many times more out of curiosity to see if it'd give any snake answers. While it does seem to parse the sentence correctly, the logic in its output is often incoherent: A: (A headless, tailless, brown creatur
195.
by blixt 6y ago
Hehe I guess this is subjective, but I tend to write mostly in American English, and I tend to use the "modernized" forms of words like monologue, hence monolog. As far as I know most dictionaries recognize it as a valid version o
196.
by blixt 6y ago
Yup this is for sure not all the outputs, but these are all in sequence, so it's able to output this quality consistently. It's just really good at generating human-like output. I posted a link to a Gist with full playground file
197.
by blixt 6y ago
Yeah it's very capable to follow patterns. Here are some more fun examples that I posted on my Twitter: https://twitter.com/blixt/status/1285274259170955265 The fact that it can take vague English and turn it
198.
by blixt 6y ago
I only wrote the "Q: ..." part – the answer, including monolog, is output generated by GPT-3. The way it works is you first give it a few sample lines, and those do include both the Q and A parts, and then GPT-3 generates consiste
199.
by blixt 6y ago
These examples I gave were on a pretty high temperature (which I think makes the fact that its answers are correct even more interesting), so YMMV. Here's a gist with the full transcript (note that several of the items at the top are w
200.
by blixt 6y ago
Since GPT-3 transforms sentences into high-dimensional vectors with connections between all the words and of course its internal weights trained on both the English language and all languages in general, I find that it has a basically flawl
201.
by blixt 6y ago
A sidenote on that last question – it's a riddle and the correct answer is a cent, but I found the output from GPT-3 quite interesting nonetheless. I'm not sure if GPT-3 is inherently able to make the connections necessary to solv
202.
by blixt 6y ago
I found that getting GPT-3 to add its own "internal monolog" in parentheses to be a helpful strategy to improve its probability of giving a more accurate answer. Here's a bunch of examples (all the answers/monologs below
203.
by blixt 7y ago
Keep in mind that this is taken with zoom lenses and that you're seeing a lot of atmosphere between the camera and the camera. This reduces contrast a lot and in an attempt to make out more detail in editing it can end up looking a lit
204.
by blixt 8y ago
They're definitely not magic, but they're also not just friction on top of regular procedures. They have use cases and used correctly they make a game mechanics into composable and configurable puzzle pieces that are easy to move
205.
by blixt 8y ago
I'm sure there are plenty of people out there who have a knee-jerk "ECS can solve that" reaction, but also be careful that you're not yourself just having a knee-jerk reaction here. A full-fledged ECS can in fact be very
206.
by blixt 8y ago
So we have 80 years to solve the issue of supporting the version going from 99.X to 2100.1 :)
207.
by blixt 8y ago
Has anyone considered using this or something else for IPC – that is, communicating between memory boundaries in the same application (e.g., Node/web, Swift or C# to a web container, etc.)? Even just for Electron the built in IPC API d
208.
by blixt 8y ago
Git also does internal housekeeping in the background when you pull. Actually most developer tools perform operations on their caches and other internal stores over time without asking or telling you about it, so this shouldn't really
209.
by blixt 8y ago
These emergent behaviors from relatively simple rules fascinate me. Have you seen https://pixling.world ? It’s basically a tool to build systems with rules for both a grid (environments) and agents on that grid (pixlings). It’s
210.
by blixt 8y ago
2FA prevents people who successfully phished you (e.g., they managed to send you an email that opened a fake Google login page that you put your password into) from logging in with your username and password without being in your physical v

More ›