phj
255 posts
I think both sides are guilty of it to some extent. Biden admin restricting Nvidia exports was probably a bad move in the big picture and democrats restricting data centres over water concerns are questionable to say the least. But not wanting the best talent in the world to work on Ai for you is troll.
Theo - t3.gg@theo
It is crazy to me that conservatives are suddenly okay with losing the global economy to China
English
@sushirollbt @itsobinnasworld @theo How low tho? At some point you lower GDP enough it hurts all Americans
English
@itsobinnasworld @theo I will happily take a lower GDP if it makes like better for Americans.
English
If you’re curious why more people aren’t speaking up about the green card news, it’s probably this.
English
im flash model homeless. 😭
Gemini 3 flash and Grok 4.1 fast got euthanized in the same week. Deepseek Flash 4 is great, but so unreliable above ~10 workers with lots of times out. Where am i to go for smart/cheap/fast LLMs that can handle 100s of concurrent workers on large datasets?
English
And one more thing to answer the question of "what are you doing with all these things", the answer is my job lol
Most of it isn't public facing, I need to start open sourcing a ton of stuff and making more of these tools public.
All the hermes agent stuff is for helping do my main job: running 3 yt channels and helping run a creator agency.
90% of the work for content is stuff u don't see. The recording it is the fun part, so is the editing. The really hard part is:
- working with 50+ brands
- dealing with 5+ inboxes
- slack
- research
- scheduling
- meetings
- and more
The hermes agent stuff is incredible for helping with that. It's also really nice for dev work. Haven't done all of these yet, but stuff like:
- auto closing issues when a pr lands that addresses them
- auto making prs when bugs get reported or noticed in posthog/sentry
- keeping track of where a project is going, quickly throwing together ideas, etc.
For the app side I've been working on:
- btca web (I desperately need to release this one, the new version is great I just do not want to deal with the stripe and db migrations so I'm procrastinating I'm sry)
- picthing
- lawn
- tons of internal tools for the channels/agency
- personal sites/resources
It'll all make it's way out over time, I have a ton to show/talk about with these. Just trying to figure out what this system is even going to look like.
Ben Davis@davis7
It's a meme but I really have been letting the psychosis take over as much as possible to figure out what I can actually do with these things - codex desktop app computer use spam - hermes agent that I've been loving - going as hard as possible on parallelizing in projects - making as many of my tools into markdown files as humanly possible - cloud agents + "cloud" agents that are t3 code instances on my mac mini - letting codex entirely control and setup my computer exactly how I want it to - unironically using gbrain, it's quite nice lol - seeing how far I can push local models on my 5090 (qwen 3.6 is very good) doing the real network setup was expensive, difficult, but was a great decicion. I feel so much more comfortable going harder with this stuff having A) a real firewall and B) hard gates around the mac mini and NAS so in the worst case scenario they can't compromise anything else on the network (the firewalla was an incredible purchase) and I still don't think I'm going hard enough
English
Forgot it was Payout Friday, had some weird bug with "minimum earnings" for some reason but still, damn. Slight dip but still holding in the $5k range
English
@paulg @Noahpinion Anywhere but Europe I suppose. Europe has other issues.
English
@Noahpinion Only in the US. It will accelerate the development of AI in the rest of the world.
English
Imagine if they had made the CLI command AGI
Leon Lin@LexnLin
why did they even name it agy and not agv in first place??
English
@theo @0xAllen_ @Dimillian Would be interesting to know if this transfers to web as in macos vs others. Are there more t3 chat macOS users on the max plan as a percentage of their respective population.
English
@0xAllen_ @Dimillian iOS users spend around 10x more on digital goods 🤷
English
Crazy how the web is still the only platform that doesn’t suck to develop for.
English
English
It’s still way too slow and generates an insane amount on tokens compared to other models on that curve. Grok 4.3 is a better choice.
Artificial Analysis@ArtificialAnlys
DeepSeek has made its temporary 75% price cut on the first-party V4 Pro API permanent, putting V4 Pro on the Pareto frontier of Intelligence Index vs Cost to Run Intelligence Index alongside V4 Flash @deepseek_ai's first-party V4 Pro API is now $0.435/1M input, $0.87/1M output, and $0.0036/1M cached input, down from $1.74/1M and $3.48/1M, with a 7:2:1 (cache input/input/output) blended price of ~$0.18/1M. Running the Artificial Analysis Intelligence Index on V4 Pro (Reasoning, Max Effort) at first-party pricing costs ~$268, ~3x cheaper than Gemini 3.1 Pro Preview ($892), ~12x cheaper than GPT-5.5 (xhigh, $3,357), and ~19x cheaper than Claude Opus 4.7 ($5,117)
English
Sorry I had to be a little more based than usual with the posts today. We'll go back to our usual AI dev posting starting tomorrow 🫡
English
@thsottiaux @ziwenxu_ I’m pretty sure they are about to hit a sub milestone anyway.
English
DID @thsottiaux JUST AGREE TO RESET THE LIMITS?!
If we get resets I'm upgrading my Codex next month from $100 to $200 plan too
MARK MY WORDS
Tibo@thsottiaux
English
I regularly get a lot of insane replies on this app. Today is no different in volume.
That said - the intelligence of the average replier has somehow gone down by like 20 points.
English
I cannot believe that these people are real
Carl@StoicConsultant
@theo Yes. This is a great example of the type of person who boasts about mediocrity because you are mediocre yourself. Thank you for your service.
English
This Gary Marcus is going around insisting that nobody will debate him about AI. I shared my thoughts and got blocked.
Pic unrelated.
English