bretpiatt

4 ms·

1.
by bretpiatt 5d ago
From what I'm experiencing, we have recursive self improvement that can find local maxima, I'm not seeing really strong evidence of unguided RSI that finds the true optimal.
2.
by bretpiatt 12d ago
Running a bunch alongside APIs for suite of internal services where domain experts are using agents to setup environments for their work. Getting them to update their LLMs to a new SDK version every time we iterate API capabilities is all t
3.
by bretpiatt 24d ago
My ability to detect sarcasm is not good. From looking at funlang's profile and other comments the profile looks like a LLM generated bot. Forums with full no verification pseudonyms seem like they have a real challenge ahead. How long
4.
by bretpiatt 8mo ago
The solution exists in the financial controls world. Agent = drafter, human = approver. The challenge is very few applications are designed to allow this, Amazon's 1-click checkout is the exact opposite. Writing a proxy for each indivi
5.
by bretpiatt 9mo ago
10 Years Ago, December 11, 2015 - Introducing Open AI -- very meta: https://karpathy.ai/hncapsule/2015-12-11/index.html#article-... The company has changed and it seems the mission has as well.
6.
Linux Foundation Announces the Formation of the Agentic AI Foundation(aaif.io)
8 points by bretpiatt 9mo ago|2 comments
7.
by bretpiatt 10mo ago
Page 5, "The knowledge cutoff date for Gemini 3 Pro was January 2025." Still taking nearly a year to train and run post training safety and stability tuning. With 10x the infrastructure they could iterate much faster, I don't
8.
by bretpiatt 1y ago
I'm pretty much doing that in a containerized deployment for project I'm looking to open source soon called webbin: Technology Stack Frontend: React 18 + TypeScript + Vite + TailwindCSS Backend: Node.js + Express + Clust
9.
Anthropic's Landmark Copyright Settlement: Implications for AI Developers(ropesgray.com)
2 points by bretpiatt 1y ago|0 comments
10.
ChatPRD AI Product Manager(chatprd.ai)
1 points by bretpiatt 1y ago|0 comments
11.
by bretpiatt 1y ago
He's been a target numerous times https://en.m.wikipedia.org/wiki/Brian_Krebs#Career Thankful he's willing to continue on the mission.
12.
GPT-5 Livestream starts Aug 7 10am Pacific(x.com)
4 points by bretpiatt 1y ago|0 comments
13.
ChatGPT Gave Instructions for Murder, Self-Mutilation, and Devil Worship(theatlantic.com)
3 points by bretpiatt 1y ago|0 comments
14.
by bretpiatt 1y ago
*slower with Sonnet 3.7 on large open source code bases where the developer is a senior member of the project core team. https://metr.org/blog/2025-07-10-early-2025-ai-experienced-o... I believe we'll see the bene
15.
by bretpiatt 1y ago
English football, unlike tic-tac-toe, can be thrilling and end in a draw. Possession mix, shots on goal, and more stats are useful to determine how exciting a match was from a box score. Frankly, for me the most boring is a 2-0 win where th
16.
Ingram Micro Down(old.reddit.com)
2 points by bretpiatt 1y ago|0 comments
17.
An Unexpected Journey with Cursor and Gemini 2.5 Pro(medium.com)
2 points by bretpiatt 1y ago|0 comments
18.
Joining CNCF as Executive Director: Let's Build What's Next(cncf.io)
3 points by bretpiatt 1y ago|0 comments
19.
by bretpiatt 1y ago
I'm running terminal in one window with AI interaction and then VS Code with project on same directories so I can see via color coding updated or new files to review in the IDE. How do you interact with your projects?
20.
by bretpiatt 1y ago
We're piloting VIA Link in San Antonio, TX to add on last mile Uber style from transit stations. Link: https://www.viainfo.net/link/
21.
by bretpiatt 1y ago
The screenshot of email from DCIO is what should be getting rolled out. This is not suspicious by itself from my perspective. SCuBA is a CISA project that improves security. Source: https://www.cisa.gov/resources-tools/
22.
Deel's CEO is now in Dubai, complicating Rippling's lawsuit(techcrunch.com)
4 points by bretpiatt 1y ago|1 comments
23.
by bretpiatt 1y ago
With their endowment above $50 billion, combined with Federal plus Non-Federal sponsored revenue at 16% of operating budget, it makes sense to me they just forgo Federal funds and operate independently. If all 16% is canceled, then they&#x2
24.
Wikipedia is struggling with voracious AI bot crawlers(engadget.com)
91 points by bretpiatt 1y ago|100 comments
25.
Every Flop Counts: Scaling a 300B LLM Without Premium GPUs(arxiv.org)
117 points by bretpiatt 1y ago|9 comments
26.
by bretpiatt 2y ago
Direct link to the TikTok video: https://www.tiktok.com/@duncan77mccabe/video/743836674477690...
27.
AT&T sues Broadcom for 'breaking' VMware support extension contract(theregister.com)
26 points by bretpiatt 2y ago|2 comments
28.
by bretpiatt 2y ago
Another use with barbed wire, we ran an x.25 network across it on a very large ranch with nodes 1 mile apart. With network monitoring we could now detect breaks in the fence at a 1 mile increment to let the ranch hands know where to go and
29.
Meta starts testing user-created AI chatbots on Instagram(techcrunch.com)
4 points by bretpiatt 2y ago|0 comments
30.
by bretpiatt 2y ago
Did you try to use GPT 3.5? Our testing is it isn't great, using GPT 4 or some of the specialized trained versions of GPT 4 (there's one with good reviews called Lisp Engineer) our experience has been different. It is not replacin

More ›