![Patrick Vlaskovits Profile](https://pbs.twimg.com/profile_images/378800000413777513/b9e4e06c7fde92d8bc85c8d291834be8_x96.jpeg)
Patrick Vlaskovits
@Pv
Followers
12K
Following
11K
Statuses
45K
🤖 Cofounder & CEO of @SuperpoweredSDK, cross-platform low latency audio SDK 🔊 Acquired by @splice.
Austin, TX
Joined December 2009
@cairoasmith Actually Smith might be the closest thing to a medieval middle class. You lived in town and had your own business! But there really is something elemental and comforting about high quality meat, cheese & bread that transcends time for some reason.
0
0
1
The calm before the AI hurricane rips through all of these....
Software and technical writing tasks were at the top; fishing and forestry had the lowest AI use. Few jobs used AI across most of their tasks: only ~4% used AI for at least 75% of tasks. Moderate use is more widespread: ~36% of jobs used AI for at least 25% of their tasks.
0
0
0
@ItIsHoeMath @CBlairforRealz There are two archetypical heroes -- the Anointed One (Harry Potter, Luke, Neo) and the Average Joe who tries hard (Rocky, John McClain). There stories are actually materially different.
0
0
0
@gbufc78 @ChefGruel I'm not in resto biz -- but have observed that there is an odd type of person who lives to bully waitstaff and draw attention to oneself that way.
0
0
14
Spies Like Us is an amazing Cold War/Reagan satire.
"Bad" Movies We Love #3 Spies Like Us (1985) - A THREAD Rotten Tomatoes Score - 35% Critics be damned! Chevy Chase and Dan Aykroyd saving us from nuclear annihilation? What's not to love? Spies is NOT a bad movie! This press conference should clear this up Read On Movie Lovers 1/
0
1
1
@credealjunkie I grew up in the Bay Area -- and I still cannot believe there is an East Palo Alto Four Seasons.
1
0
1
@credealjunkie Ah. You can absolve yourself of sin by making your buyout to them especially tasty. But all that buyout must go to them vs slicing off a bit to a neighborhood org.
0
0
1
NO MORE FULL SEND YEET IT ALL INTO THE GIGANTIC CONTEXT WINDOW ANYMORE? Fml.
New research shows that LLMs don't perform well on long context. Perfect needle-in-the-haystack scores are easy—attention mechanisms can match the word. When you require 1-hop of reasoning, performance degrades quickly. This is why guaranteeing correctness for agents is hard.
0
0
0