@SignalRounds

Clinical informaticist. Evidence on healthcare AI — what the claim supports, and what it doesn’t. YouTube: Signal Rounds

Joined December 2015
I work in clinical informatics. On YouTube as Signal Rounds, I cover healthcare AI the way I’d want it briefed before a real workflow decision: • what the claim actually says • what evidence class it is • what it does not prove No hype. Not a vendor channel. Sources in the description. Start here: youtube.com/@signalrounds
3
554
On my road trip through Canada (booked with @Muse and @bot, planned with @NousResearch Hermes Agent using Astra). At the motel i am at, GrokBot called me in some food using @ChatGPT Live-1. The guy said it sounded like a beautiful woman and he had no idea it was an agent at pickup 🤣
2
2
124
I wanted to share how I am using AI in the real world. Last night, I wanted to test the capabilities of @Muse. Using my Hermes Agent running Astra, I had it plot a detailed itinerary for my drive from Alaska to the east coast. It’s a 10 day trip. I am moving and traveling with a dog and a cat so hotels are limited. Astra planned my route, but Muse was able to setup the hotels completely, and even made some positive changes during the process. The Muse app is excellent and snappy. I would say it is at least as good as @bot. Last year, it took me hours of research to do this and hours more to book all of the rooms. My agents did most of this work for me in one evening while I spent quality time with my infant. This is the future I want to live in. I believe we need to continue to accelerate. The possibilities only open up from here.
1
1
5
226
This is a guy that knows his stuff. I trust him and Nvidia more than any AI labs
open source won’t pace
37
Crazy that this is going to be the reality
2
101
Dariyan Jones retweeted
Hah! Michael changed math to cancer & wrote a similar letter to the one Fields Medalist mathematicians wrote expressing deep concern about AI solving math problems! Read this and let me know if anyone in the world would side with cancer researchers if they wrote such a letter!
17
31
4
291
24,181
The labs that already own the frontier just asked everyone else to pace it. A slowdown by OpenAI, Anthropic, and even xAI does not serve mankind. It serves the people who already have the models. The government and “select partners”. Now more than ever we need to push back. Open source is becoming more important by the day. Soon, we will hear how we need to ban any open source or models that are made from excluded countries like China for “our safety”. It’s a play to give more control to the US government and to allow our labs to ensure they stay ahead. Fight like your loved ones lives depend on it. Because it might.
Effective Accelerationism (e/acc) Now needed more than ever! We won't give up without a fight.
1
154
Dariyan Jones retweeted
As some of you know, we have a new e/acc group chat. We're the resistance to the new wave of AI doomerism. The chat is already pretty big, but I'd especially like to invite: @Sentdex @perrymetzger @MindsAI_Jack @alexwg If you'd like to join, DM me.
92
29
3
477
12,597
Dariyan Jones retweeted
The anti-AI Coxon operation is being exposed! At this point, you would have to be spectacularly naive to look at this pattern & dismiss it as ordinary, spontaneous engagement. The scale, timing, & geographic distribution of the engagement on this post, clearly suggests this was deliberately amplified in a very organized and planned manner. For a post driving an American political narrative here are some extraordinary data points: Using a sample of 3,500 reposts, Digital Borders estimates that 76% of the engagement on Coxon’s post came from outside the US, primarily from India, Indonesia, and Mexico! The deeper people dig, the stranger this gets…or maybe not so strange after all.
Another fascinating indicator that there might be some kind of astroturf campaign at work. Using a sample of 3,500 reposts, Digital Borders estimates that 76% of the engagement for Coxon's post came from outside the US, mainly India, Indonesia, and Mexico.
5
32
1
165
10,853
Grok 4.7 needs to cook a little longer
Replying to @farzyness
Grok 4.7 needs a few more days to cook. We might have penalized response length too much (or something) in RL, as it still gives up on hard tasks (that it can do!) too early and isn’t yet sufficiently rigorous in checking its work.
33
Well surprise surprise 😦
BREAKING: ex-Anthropic Doomer Jacob Coxon tied to "Newspeak", British NGO involved with "hacking" democracy, Hillary Clinton, prison abolition, and extinction rebellion
1
29
Dariyan Jones retweeted
Astra for checking scientific papers
I sicced Astra on a bunch of replication packages, and it found a huge number of coding errors. Most of these were inconsequential, but some of them overturn central results for papers in high-ranking journals. Astra also discovered that many models weren't even run properly.
96
67
7
1,607
205,100
Hey all, just woke up from a nap. Did @OpenAI release Astra to the public?
1
60
Dariyan Jones retweeted
1
1
29
2,626
Yeah it’s Thursday! Astra 🚀
OpenAI is holding a live keynote for businesses on "how frontier AI is reshaping cybersecurity", featuring co-founder and President Greg Brockman... this Thursday at 1pm PT. OpenAI's usual launch time is 10am PT. Not suspicious at all!
74
I am so glad I jumped on this offer when it came out. Yeah Grok 4.3 was lackluster, but don’t bet against Elon. Now it’s paying off with Grok Build, Grok Bot, and Cursor Ultra all included.
1
59
It is inevitable
SpaceX is going to win the AI race. Believe.
1
6
646
This is why you can't keep letting them break you down and keep chipping away. Eventually they have your soul.
“No one is aborting babies at 9 months” -The Democrats Massachusetts: “YASSSS WE MADE IT LEGAL TO ABORT BABIES AT 9 MONTHS! WE LOVE THIS JOURNEY FOR US!!!”
67