dazzaji

Nope. Kept them coming!

I’ve been using Sublime Text many times a day since around 2010 for my daily and weekly logs. My Mac is set to use it as the default for .py, .md, and .txt files, so I find myself in Sublime all the time—and I absolutely love it! Sure, I develop in VS Code, but I live in sublime. It’s fast, clean, and dependable. Thanks to the team for building such a fantastic editor!

Another thought: Focusing resources on promoting creation of the most powerful open source models rather than continuing the overwrought emphasis on risks and safety is a vastly wiser course. The R1 models recently out of China are a solid case in point. The real risks will follow from failing to compete, and the real rewards will follow from succeeding. you can’t regulate and restrict your way into successful competition. In fact, it is quite the opposite.

China is leveraging open-source AI models to gain global influence, potentially creating dependencies on its technology and values. This article argues that U.S. policymakers are focusing on the wrong threats by restricting open AI models and suggests that embracing open-source innovation is key to maintaining U.S. leadership in AI. It outlines strategies like investing in open development, promoting adoption, and avoiding restrictive policies to counter China's growing AI dominance. A compelling call for 'open-source diplomacy' in the AI race. This was written over a week ago, and now with the advent of the R1 model out of China it seems even more relevant and timely.

This guide compares top general-purpose AI models like ChatGPT, Claude, and Gemini, highlighting recent advancements, strengths, and use cases. It offers practical advice on choosing the right AI for various needs, making it I think a useful resource for anyone navigating the fast-changing AI landscape.

Honestly, LLM-related posts are usually a topic I’m more likely to click on, depending on the article. I find some of the discussions pretty interesting and useful, though I get why the saturation might feel like too much for others.

This article does a great job summarizing the rapid advancements in reasoning and AI agents. Models like OpenAI's o1/o3 and DeepSeek's r1 demonstrate how inference-time compute and structured Chain of Thought (CoT) are pushing LLM capabilities in STEM and coding tasks. The speculation about pivot words and backtracking behavior learned through reinforcement learning is particularly intriguing—it could be transformative for reasoning in domains with external verification.

The discussion on agents and moving beyond chat interfaces toward workflows like Cursor resonated with me. A shift in Human-AI interaction paradigms feels essential to unlock the full potential of autonomous agents. However, as the author notes, error rates and cost remain significant hurdles.

I've been experimenting with multi-agent systems in Python for the last year and find measuring performance and success one of the hardest parts. While today's LLM agents are still primitive, they already show immense potential. Even without advances in base models, creative agent design patterns could unlock more functionality, and with better reasoning and larger context windows, the possibilities expand even further.

Thanks for clarifying that timescape cosmology focuses on dark energy, not dark matter—makes sense they’re addressing different puzzles. That said, I’m curious: could there be any indirect connection between the two? For instance, if dark matter’s influence on gravity shapes the 'lumpiness' of the universe, could that, in turn, amplify the time dilation effects discussed in timescapes? I realize the authors didn’t delve into dark matter, but wondering if anyone has explored how these two phenomena might interact. Definitely out of my depth here, but this potential connection is what sprung to mind and on rare occasions my ill-informed intuitive curiosities turn out to have unexpected merit, so I’m asking if there could be anything to it?

Just a thought – could the existence of something like dark matter actually be causing the differences in the rate of time in different parts of the universe? If dark matter affects gravity in ways we don't fully understand, maybe it's also influencing time dilation across the cosmic web. That might reconcile these two ideas – with dark matter being the underlying cause of the time variations that timescape cosmology talks about. I'm curious if anyone has explored this connection or if it's completely off base?

Come to think of it, another angle to consider – could it work the other way around? Instead of dark matter causing time dilation, could time dilation itself be a root cause that leads to the accumulation or creation of something like dark matter? If time moves differently in certain regions, maybe it affects how matter and energy interact over cosmic timescales, creating the conditions we interpret as dark matter. Just wondering if this has been explored or if it’s way out there as an idea? I’m definitely not a cosmologist!

I’m now free-falling down a moon viewing ceremony rabbit hole of emoji history, and enjoying the ride!

I use the Apple Vision Pro pretty much every workday for at least one to three hours. Sometimes, when I’m deep in coding or other major projects, I’ve been in it for over five or more hours at a stretch. I use it exclusively with my MacBook Pro or iMac as an immersive display, with the Magic Keyboard and Trackpad paired to the computer. This setup is basically perfect. I frequently have some mellow ambient music in the background, which, in combination with the immersive work environment, helps me concentrate like never before for long stretches without getting distracted.

A major improvement happened with the release of beta version 2 of their operating system, which made the keyboard visible even when the immersive environment is fully turned up. This change allowed me to have the best of both worlds. Obviously, I have it plugged into an outlet through the battery or it would never last for such a long time. I’ve tried going back to the Oculus, but there’s no comparison. I love it!

Thanks for clarifying, and I’m sorry to hear this is happening. LLM agents have a lot of promise, and it’s frustrating to see baseless fear being stirred up. There’s already enough uncertainty around what’s legitimately needed to get the tech and usage right.

I saw that but here’s an alternative take on what happened:

While the session file definitely shows the AI agent using sudo, these commands were executed with the presumption that the user session already had sudo privileges. There is no indication that the agent escalated its privileges on its own; rather, it used existing permissions that the user (buck) already had access to.

The sudo usage here is consistent with executing commands that require elevated privileges, but it doesn’t demonstrate any unauthorized or unexpected privilege escalation or a self-promotion to sysadmin. It relied on the user’s permissions and would have required the user’s password if prompted.

So he sudo commands executed successfully without any visible prompt for a password, which suggests one of the following scenarios:

1. The session was started by a user with sudo privileges (buck), allowing the agent to run sudo commands without requiring additional authentication.

2. The password may have been provided earlier in the session (before the captured commands), and the session is still within the sudo timeout window, meaning no re-authentication was needed.

3. Or maybe the sudoers file on this system was configured to allow passwordless sudo for the user buck, making it unnecessary to re-enter the password (I just discovered this one, actually!).

In any case, the key point is that the session already had the required privileges to run these commands, and no evidence suggests that the AI agent autonomously escalated its privileges.

Is this take reasonable or am I really missing something big?

Wait, is that gist of the same session as is described in the article? I don’t see any escalation of privileges happening.

Among the good reasons for SB-1047 to have been vetoed are that it would have regulated the wrong thing. Here’s a great statement of this basic flaw: https://law.mit.edu/pub/regulatesystemsnotmodels

Not speaking for MIT here, but that bill needs a veto and a deep redraft.

For Google Docs like real time collab on native markdown, I like and use daily:

* https://hackmd.io

and

* https://stash.new

It’s about agents. This is the first clear public signal from OpenAI that they are in fact going to release agentic AI capabilities (enabling the model to take actions on behalf of the user, and this will inevitably eventually include consumer and other commercial transactions). Altman says the model will “…. the ability to take actions on your behalf”. The mention is rather buried near the bottom but it hit me as the most important and profoundly transformative (mostly in good ways but also potentially very risky ways) roadmap item yet announced.

This looks really promising for complex legal reasoning tasks and other challenges. How can I track progress? Is there an email list or something? Thanks!

Yes, and it works like a charm! I've used it a few times already today and can see this becoming a very regular part of my workflows. Thank you!

I’m getting the server overload error but assuming this mostly works I’d use it every day!

I’m way out of my depth here but if I’m reading this right a flow model would not base outputs on a prompt but instead, after it’s sampling/inverse flow pipeline (if I got that right?) is activated it would somehow just output whatever image it’s statistically sampling at the moment. Is that even close?

Go for it!

If I ever create a working definition of artificial consciousness, I’m definitely calling it quidlesmoopy.

Evidently, there is now a legislative push for parliament to reverse these convictions:

https://www.bbc.co.uk/programmes/p0h48cft

“UK to bring in law to quash wrongful Post Office convictions”

According to the BBC Global News Podcast “The legal step followed public outrage after a TV”

I’m intrigued and currently downloading this app. Love the idea of having offline direct access to this model. One small-ish thing though: Looks like the URL for the privacy policy (http://opusnoma.com/privacy) linked from the App Store page goes nowhere. Actually, opusnoma.com is likewise offline.

I meant “first test” not first text. Though it was the first text it output so that too, I guess.

I love this so much. I’ve got a backlog of videos to comb through and this might just be the solution. To me, immediate usefulness is a great informal indicator of a righteous hack. As a first text I threw a rather dense talk I gave to Legal Hackers DC chapter early last year at it and got a very solid set of summaries [1] broken down into chapters that I didn’t even realize existed but in retrospect totally agree with. And the most convenient part is that you support static links to the summaries so I can reshare it. One small question that hit me immediately is the copyright all rights reserved at the bottom of each summary. I’m not necessarily poking criticism of that but do wonder if it’s really what you want? Especially with the built in share feature it seems like at least some rights to reuse and such are expected. Anyway, overall this is stellar work, easy to use, and immediately helpful. Thank you!!

[1] here’s the summary I was describing: https://www.videogist.co/videos/generative-ai-for-law-1021

Does anybody know if 2008-2009 SAT is in the training set for these models? Assuming so, I’d be especially interested in head-to-head evals on this type of non-code benchmark for problem sets not already in the training data, to see how it performs on fresh situations.

Excellent! And you were just saying how risky it is to rely long-term on OpenAI text embeddings in your post on the topic. The timing for this open source option worked out nicely.