atlex2

8 ms·

31.
by atlex2 2y ago
I continue to prefer Claude over ChatGPT when it comes to discussing matters of human-human interactions. Opus tends to understand the subtleties of social interaction better than 4 and definitely 4o in my experience.
32.
by atlex2 2y ago
I have an even simpler prompt for LLMs that run on mobile: "how many feet are in four feet"-- confabulations on this one sometimes revert to counting toes.
33.
by atlex2 2y ago
Hello! This sounds interesting, but I wasn't immediately understanding it. Does your project enable models like Llama, which don't have native tool use, to have tool use? If so, that would be super useful!
34.
by atlex2 3y ago
Brian is named the grand integrator of 2024, probably one of the fastest on the planet.
35.
by atlex2 3y ago
Update: hmm. I might be mistaken. There are articles that measure insulin levels after drinking the soda. These show no connection for diet soda. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5118762/ https:&
36.
by atlex2 3y ago
Diet soda and sweeteners behind them are still a scam even if they don’t “cause cancer.” They’re a scam because they still cause insulin response. Pancreas cranks out insulin after tasting the soda, and there no blood sugar to react with. M
37.
by atlex2 3y ago
For things that fly, what I’ve seen be more significant is the weight savings (as long as you’re not running too rich).
38.
by atlex2 3y ago
The 3.5 prompt/model should probably be updated to inform users of their privacy more accurately. > Me: is our conversation confidential? > 3.5: As an AI language model, I do not have the ability to store or share any information
39.
by atlex2 3y ago
I’ve gotten around this when I need by deleting the viewport containing the headers and moving out the large chat div, then I use one of the debug screen cap commands built into chrome on that div to save a giant tall image. Kinda bad, they
40.
by atlex2 3y ago
Pretty sweet. Is it small enough to run on an iPhone?
41.
by atlex2 4y ago
Forensic Architecture features regularly on the front page of HN, such as in `Beirut Port Explosion.` They also have a major presence in the art world. This article describes how FA evades fact checking by positioning their work as art and
42.
Forensic Architecture: Precise not accurate(artnews.com)
1 points by atlex2 4y ago|1 comments
43.
by atlex2 4y ago
Confused why you can’t use this to transcribe from any AirPods in your vicinity? I thought anyone could subscribe to a btle gatt attribute.
44.
by atlex2 5y ago
Winter is almost over, that helps a lot.
45.
by atlex2 5y ago
Pilot here. That aircraft wasn’t built with an electrical system, and people only install them when they need to fly into populated areas— it’s basically an antique. No way to radio anybody, barely any checklists to follow; generally extrao
46.
Eric Schmidt Talks 'What He Needs' from MIT CSAIL(youtube.com)
3 points by atlex2 9y ago|0 comments
47.
by atlex2 12y ago
There’s a lot of discourse on whether to “drop out” of college to work on a company. I think current conversations are important, but misdirected.  Namely, the question “Should I dropout?” misrepresents the end people are striving to achiev
48.
Do you really need to eat your vegetables? Deliberating dropping out MIT.(medium.com)
2 points by atlex2 12y ago|1 comments
49.
by atlex2 13y ago
This is a great step-- I note that you chose to open source after your were on the store. As a well-known company, was that to prevent pre-release copycats?
50.
by atlex2 13y ago
Works skippingly for identifying users' reasons for app failures.
51.
by atlex2 14y ago
Thanks! For sure drop us a line and we can get you building on it! everyone@mosaic.io
52.
by atlex2 14y ago
Using some of this code, I think screen-stitching can finally be easily used by devs in mainstream apps. We actually wrote unit tests during a hackathon. :)