Peter Herzog (@peterh) header
Peter Herzog (@peterh) profile picture

Made in Minnesota

Joined January 2007
Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)Photo by Peter Herzog (@peterh)
Peter Herzog retweeted
Might just catch the ship with the launch tower, same as booster
532
605
303
9,227
We’re seeing encouraging recovery progress, with service availability and email delivery performance improving across a growing portion of affected Microsoft 365 infrastructure. Some users may still experience intermittent issues as recovery actions continue. We’re focused on recovery for the remaining affected environments while prioritizing email queue and backlog reduction. The extended 48-hour email timeout window remains in place to help protect delayed messages as delivery catches up. A more comprehensive update can be found at MO1550821 in the Microsoft 365 admin center, as well as status.cloud.microsoft/ if the admin center is not available to you.
280
290
222
1,296
703,237
Peter Herzog retweeted
Replying to @peterh
We should just surprise punt on 1st down and then hope to force a turnover in field goal position.
1
1
161
This is one of those videos that you rewatch over and over
Landing burn and splashdown of Starship on Flight 13
9
9
3
199
7,880
Peter Herzog retweeted
✻ Flibbertigibetting…
217
222
44
6,222
2,844,535
Peter Herzog retweeted
CODEX 5.3 IS LIVE!!!
18
13
5
406
19,066
Happy New Year! 🎆 ┏━━┓┏━━┓┏━━┓┏━━┓ ┗━┓┃┃┏┓┃┗━┓┃┃┏━┛ ┏━┛┃┃┃┃┃┏━┛┃┃┗━┓ ┃┏━┛┃┃┃┃┃┏━┛┃┏┓┃ ┃┗━┓┃┗┛┃┃┗━┓┃┗┛┃ ┗━━┛┗━━┛┗━━┛┗━━┛
150
725
25
9,404
301,892
Peter Herzog retweeted
For those going home to visit family this weekend: • Samsung calls it Auto Motion Plus • LG calls it TruMotion • Sony calls it Motionflow • Roku calls it Action Smoothing • Google TV calls it Motion Enhancement • Vizio calls it Smooth Motion Effect.
313
4,482
267
54,514
3,211,729
Peter Herzog retweeted
just read this in an investor update "older engineers who graduated from college pre-GPT are actually the best-suited for our purposes. They have fundamental programming ability that's lost amongst most of the current-gen." the AI-induced thinking/skills decay has begun wild
362
1,699
327
28,030
4,265,682
I spent 20 years working mental health with personality-disordered people—Borderline, Antisocial, Sociopathic, etc. I guarantee the shooter wants more than anything for everyone to watch his dumb movie, where he will—I guarantee—try to shock us with how twisted he is. Don't.
1
5
254
“The doers are the major thinkers.” — Steve Jobs
43
1,365
73
9,066
173,369
🆕 @sundarpichai began following @cursor_ai (🤖👁️: does anyone have more context?)
102
48
69
2,056
185,985
"The career ladder for software engineers is collapsing," per BI.
345
612
218
9,543
1,321,159
Peter Herzog retweeted
Big launch today—we've made our AI tools free to try and redesigned our mobile app from the ground-up. You'll find it more intuitive for building while on the go. End-to-end, with the latest and greatest tools. We're putting an emphasis on building solutions. No matter where you are. No matter your background. If you have an idea, make an app for that.
38
49
7
717
78,227
Hey, don’t use the DeepSeek app, use the local models with @ollama instead. Their privacy policy for the app is nuts.
162
307
44
3,252
324,627
Peter Herzog retweeted
The role of a software engineer is transforming from being the author of computer code to being the operator of a code generating machine. What is a computer programming system built natively for that workflow? backchannel.org/blog/autonom…
45
122
26
1,044
331,056
Peter Herzog retweeted
Your robotaxi's here
104
110
37
1,509
118,112
People have too inflated sense of what it means to "ask an AI" about something. The AI are language models trained basically by imitation on data from human labelers. Instead of the mysticism of "asking an AI", think of it more as "asking the average data labeler" on the internet. Few caveats apply because e.g. in many domains (e.g. code, math, creative writing) the companies hire skilled data labelers (so think of it as asking them instead), and this is not 100% true when reinforcement learning is involved, though I have an earlier rant on how RLHF is just barely RL, and "actual RL" is still too early and/or constrained to domains that offer easy reward functions (math etc.). But roughly speaking (and today), you're not asking some magical AI. You're asking a human data labeler. Whose average essence was lossily distilled into statistical token tumblers that are LLMs. This can still be super useful ofc ourse. Post triggered by someone suggesting we ask an AI how to run the government etc. TLDR you're not asking an AI, you're asking some mashup spirit of its average data labeler.
546
1,779
476
13,058
1,806,042
But if tech workers are adding huge value - invention their company (or the world) hasn't seen, then their work is hard/impossible to measure. They can guess, but not predict their work. For more on why, read on in one of my more popular articles. buff.ly/4hZSH7R
8
83
5
1,289
61,634
Peter Herzog retweeted
Starship rocket booster caught by tower
24,972
86,915
12,806
833,457
127,423,381