Jamie Gray (@jamiegraytech) header
Jamie Gray (@jamiegraytech) profile picture

Unpace the Frontier

USA
Joined April 2015
Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)Photo by Jamie Gray (@jamiegraytech)
“If machines produce everything we need, the outcome will depend on how things are distributed” - Stephen Hawking, 2017
7
15
3
118
another reason to switch to Linux
1
79
Jamie Gray retweeted
good prompt "find all of the implicit state machines in my app"
165
396
286
8,446
2,141,018
"western values" such as "the people overthrowing a tyrannical oligarchy" is gonna look extremely Maoist when you draw it all the way out ✅🤓
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values. Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line. After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs. The values in an open model are one fine-tune away. Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
71
Jamie Gray retweeted
Is there a mister alignment
37
110
12
1,616
41,165
You still click yes? - -dangerously-skip-permissions
Me clicking Yes to all to claude code without reading anything
108
Jamie Gray retweeted
Calling Vibe Coding "slop" while having a career built entirely on copy pasting from Stack Overflow is peak irony.
99
300
34
5,475
70,838
Linus Torvalds says, rules prohibiting AI usage are “UTTERLY STUPID” and “idiotic”. For the Linux kernel, Torvalds says, “we’re not going to make stupid rules that say ‘you can’t use tools for X’”.
162
812
255
7,372
503,966
Jamie Gray retweeted
the lion does not concern himself with ESD protection
68
351
5
9,570
69,923
You can now book our humanoid cleaning service in Los Angeles at tau-robotics.com. If you don’t have an invite code yet, join the waitlist. Watch our humanoids cleaning customers’ homes in the thread below.
131
114
69
1,094
162,606
Jamie Gray retweeted
95% of code was slop before AI existed
160
135
29
2,088
35,548
Jamie Gray retweeted
BREAKING: A hacker accessed Denmark's national ID database with the password "123456," exposing details of 8.8 million people
945
3,037
1,462
27,051
2,439,684
I have lost so many Obsidian notes from iCloud sync colliding with Obsidian Sync Nightmare
1
138
Jamie Gray retweeted
Fyi this is the guy that got caught. Probably bought this shit with a credit card and delivered to his address. Alibaba ass spy
Replying to @OpenAINewsroom
the noob used seeed studio, second most newbie thing than arduino uno give him extra 10 years for that blob on the solder
39
13
3
634
27,862
Post-scarcity price equilibrium is zero btw
70
circuit cardigan
221
3,200
615
41,851
1,275,436
Jamie Gray retweeted
openclaw was this year btw
2026 has been about 5 years so far
217
695
111
19,182
340,855
Jamie Gray retweeted
😭 roon why
13
7
1,289
32,625
Jamie Gray retweeted
dear people who say ai generates bad code, pls show me the handcrafted code you were shipping before ai best regards, can
142
102
22
2,368
61,430
Closed source is dead. REA lets Claude Code or Codex take apart any app, all the way down to the binary, and explain exactly how a feature works. - Native apps, Electron apps, websites, APKs, firmware, even smart contract bytecode - Shows the evidence for every claim - Then builds its own version of the feature 25,000+ GitHub stars in a single day. 100% open-source. github.com/morluto/rea
170
716
145
7,916
946,583