Shows custom handmade video of product, final image is cheap plastics shit!

Also people are sleeping on SWE-2 from @cognition it’s good, although their billing and plans, limits need attention, let’s see what happens after they charge for it.

Anyone complaining about Astra & OA models have a skill issue. suggestion get the ai to write your goals.

hot take: unused weekly Al limits should roll over to the next week. if l use only 50% of my Claude or Codex limit this week, I should start next week with 150%.

Worse than Kimi K3, GPT-6, and GLM-5.3. GPT-6 is the smartest, Kimi K3 is the most tasteful, and GLM-5.3 is the most trustworthy at the terminal. In my limited testing so far MiMo seems to struggle to follow instructions.

I got a full reset for the entire month back to 100%

Fuck AI monthly & weekly limits

Silcone valley realised that they are putting there software engineers & saas investments out of business and are now charging fair comparable market rates.

@thsottiaux We are grateful for the resets, but I tried the auto billing expecting to get some leg room out of the credits with a 40% discount, but yeah just burnt through $, seems off compared to normal usage. The workload was much less in my opinion with credits. Astra medium.

Looking at the dashboard we might hit a new milestone to celebrate tomorrow. Hold on to your Codex

The irony of AI: Smarter than a math PhD. Solves 10 open math problems. Dumber than my intern. Can’t follow 3 simple instructions, but somehow burns 10M tokens and $250.

Cue the vibe coders crying about maintenance!

I’m begging you, please don’t let AI write your content from scratch. Everyone knows you didn't write it, and it has zero personality. You are effectively a bot. Use AI as an editor, not a ghostwriter

RT @tunguz: Surround yourself with people who challenge you to become a better version of yourself. Run away as far as you can from people…

One word: idiots

Yet we still don't have level-5 self-driving cars, and certainly not self-serving cars that can learn to drive in a few hours of practice like any teenager. We don't even have domestic robots that can do what 10-year olds can do the first time we ask them. We don't even have robots that are nearly as smart as a house cat. The G in AGI is nonsense.

So the market fears that AI labs like OpenAI and Anthropic will render $PLTR's enterprise software obsolete. When asked if Anthropic is a competitor, Karp said that most of the capabilities frontier models discuss in public are currently running on Palantir. At its current scale, Palantir acts as the underlying infra for these AI labs, functioning as a 'nation state' in the ecosystem. Unlike LLMs, Karp says high-end enterprises such as defense, aerospace, manufacturing require deterministic execution where AI errors are catastrophic. Commercial enterprises are becoming disillusioned with frontier AI labs because raw models lack the capacity to handle strict security, data integration, and ontology, leaving them almost useless for complex corporate infra.

It seems almost too dumb to be true, but apparently the literal belief of SaaS bears is "all software is a 0 because Claude can one-shot these apps" Just staggering levels of short-sightedness in that statement

Sounds like the least plausible propaganda I’ve heard since that time when ⁠Kim Jong-Il tried golf for the first time and finished with 11 holes-in-one to achieve a 38-under-par game on a championship 18-hole golf course. Reports say each of his 17 bodyguards verified the record-breaking feat. He then decided to retire from the sport forever.

seriously be careful out there everyone i had 2 Microsoft Copilot licenses in my car, and someone broke in and left 4 more