@jetsetter@mastodon.cloud (@jetsetter) header
@jetsetter@mastodon.cloud (@jetsetter) profile picture
Joined October 2006
Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)Photo by @jetsetter@mastodon.cloud (@jetsetter)
"Reflow" should be considered for the word of the year @MerriamWebster
134
Gemini Project Astra AI bike demo is wrong step one to fixing the bike is to recommend the person toss entire bicycle the bin because it is a Huffy. #GoogleIO
1
202
Had a submission on Hacker News hit #1 today, holding a top five ranking for over 7 hours: news.ycombinator.com/item?id… hnrankings.info/43902263/
119
Any evidence of widespread internet infrastructure problems / issues in Portland / west coast? Seeing anecdotal evidene of Verizon wireless internet sending iphone into SOS mode, Comcast issues downtown and Quantum fiber issues in SE.
2
770
Just shipped a major update to FileKitty! Now with history tracking — step back through past file selections & outputs. FileKitty is a simple macOS tool for selecting, combining, and copying file content — perfect for giving LLMs real context. 🔗 github.com/banagale/FileKitt…
74
chatgpt needs some kind of paddle shifter UI and hotkeys flip between GPT-4o and o1-preview. -> Kick off a very hard code analysis in o1 then work with the results in 4o. When ready, return to 1o with the context created by the 4o exchanges.
1
183
Right now, you have to flip between the llm products using the dropdown or stop and resubmit which are both clunky.
1
106
o1’s feedback loops are too slow for software dev. My experience with o1 is the quality does not make up for the exceptionally slow speed. 4o is great in part because it it is so fast. With the right prompting it delivers competitive results.
1
89
This guy cackling through the man burn live stream.
78
Just gave eight consecutive demos of Vision Pro on VisionOS 2.0 beta 3. AMA.
156
New ChatGPT application for MacOS puts OpenAI one modifier key awat from competing w Spotlight search! Spotlight's held persistently low value: clunky in caching, showing things you aren't looking for--just not great. Apple Intelligence can't come soon enough.
1
131
The styling of the two options is so similar, it must be purposeful. Pull up Apple's native Spotlight search: Command-Space Pull up OpenAI's portal to GPT: Option-Space 😬😬😬
1
91
One thing Dall-E / gpt-4o are still absolutely terrible at is UI generation. This series on an increasingly clarified request for a modal for adding to and editing a table of abbreviations. What is that stylus even doing in there?
127
X is not a very good name for this product. Or really, any product.
1
94
“If you know what the future holds—what it will look like—you can buy it cheap, today.” Crook Manifesto by Colson Whitehead
1
94
If reports are correct, this might explain why Apple has not been able to settle with Gemini.
Introducing GPT-4o, our new model which can reason across text, audio, and video in real time. It's extremely versatile, fun to play with, and is a step towards a much more natural form of human-computer interaction (and even human-computer-computer interaction):
1
166
Top ten, front page HN, baby.
I built a little macOS utility app to aid coding with ChatGPT and other LLMs: github.com/banagale/FileKitt… Select multiple source code files and FileKitty concatenates them and labels them with their shared filepath in markdown. #opensource #chatgpt #LLM #OpenAI
1
1
561
I built a little macOS utility app to aid coding with ChatGPT and other LLMs: github.com/banagale/FileKitt… Select multiple source code files and FileKitty concatenates them and labels them with their shared filepath in markdown. #opensource #chatgpt #LLM #OpenAI
1
357