That’s hilarious if it’s real

U.S. Automakers Risk Being Reduced to Niche Producers of Gas Vehicles

U.S. Automakers Risk Being Reduced to Niche Producers of Gas Vehicles www.nytimes.com/2026/03/03/b...
Well Boohoo automakers. You’ve hitched your wagons to Ratpublican politicians and they’re screwing you. As a company you’ve screwed your employees, you drive cost down buy building unreliable cars.

Do IDEs do RAG across previous chat sessions in the same project? I think that would be helpful…

someone at the pentagon frantically typing “Claude, open the strait of Hormuz for me, quickest possible strategy, make no mistakes.”

Has anyone setup ai as an out of office responder that can still do small tasks?

EVs are still only ~5% of the global car fleet.

Yet already displacing ~1.5 million barrels of oil per day — the early stage of structural demand erosion.

Disruption doesn’t start when something is big, but when the curve begins to bend.

The #Bettrification S-curve has crossed that point.

I’m not liking the trend of recent ai coding tools running processes in hidden terminals. I like to be able to see what’s running and easily stop it or re-run it

From the singularity community on Reddit: OpenAI researchers hinting at an omnimodal model coming Explore this post and more from the singularity community

The era of the world model driven LLM might be upon us

Imagine the reasoning trace being an actual video of the model simulating some process, in order to derive an answer…

From the StableDiffusion community on Reddit: Just compiled FP8 Quant Scaled of LTX 2.3 Distilled and working amazing - no LoRA - first try. 25 second video, 601 frames, Text-to-Video - sound was 1:1 ... Explore this post and more from the StableDiffusion community

Pretty cool comparison of fp8 vs bf16 on latest Ltx model

We’re going to be doing local world models soon… going to be huge for the robotics industry

Might seem like we’re in an LLM lull now but we’re about to see an explosion of new capabilities later this year

If a being is intentionally created, maintained, and revised within an artificial design framework, then modification may be part of the normal moral relationship to that being, not automatically a violation of it

From the singularity community on Reddit: Pentagon Refuses to Say If AI Was Used to Select Elementary School as Bombing Target Explore this post and more from the singularity community

I think this is where the attack on Anthropic is actually coming from

They don’t want Anthropic to blow the whistle so they’re leaning on them

someone did a more rigorous analysis of this and confirmed, yes, Qwen is slightly but consistently better than GPT-4o

x.com/n8programs/s...

Caitlin KALINOWSKI over X I resigned from OpenAl. I care deeply about the Robotics team and the work we built together. This wasn't an easy call. Al has an important role in national security. But surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got. This was about principle, not people. I have deep respect for Sam and the team, and I'm proud of what we built together.

OpenAI head of robotics just resigned over company deal with the Pentagon saying…

“Surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got”

Today, Michelle and I are proud to announce that we will be hosting the dedication ceremony for the Obama Presidential Center on June 18th in Chicago, and welcoming the public on June 19th.

We can’t wait for you to visit. Go to obama.org to learn more.

From the ChatGPT community on Reddit Explore this post and more from the ChatGPT community

This is pretty amazing. A user uploaded a DLL file of a core game logic to the ChatGPT web ui and told it to make some modifications

It’s unclear exactly what it did but my suspicion is it used a hex editor to analyze the dll and patch it

More proof we aren’t pushing ai hard enough yet

Very interesting. Myself and others have been saying for a while that ai will be the only real UI in the future, and apps will just be ways of connecting ai to data sources

Prediction is even more true today

Cool to see the new Rivr robot. They were prototyping based off of Swiss mile dogs before, much more similar to the Unitree dogs we have seen so much of. But now they have custom hardware which seems to make much more sense for package delivery.

Indonesia to Block Children Under 16 From Social Media The ban is to take effect March 28, according to a government minister, but details about how it would be carried out were scarce.

Indonesia said that it would bar anyone under the age of 16 from access to social media, joining a growing list of countries that are enacting such restrictions in a bid to safeguard the well-being of children.

GPT-5.4 Pro (xhigh) also improved CritPt record from Gemini 3.1 Pro's 17% to 30%. OpenAI appears to have an edge on the hardest math and physics reasoning tasks.

"CritPt evaluates language models on solving unpublished, frontier-level physics problems that require genuine research-scale reasoning."

Wow

The image is a benchmark comparison infographic titled "Qwen3.5-4B vs GPT-4o." It compares the Qwen3.5-4B open-weight model (released March 2026) against OpenAI's GPT-4o (from May 2024). Summary of Results * Total Wins: Qwen3.5-4B wins 5 out of 7 benchmarks; GPT-4o wins 2 out of 7. * Average Advantage: Qwen has a +9.6 average advantage over GPT-4o across the categories shown. Benchmark Performance (Bar Chart) The bar chart displays percentage scores across seven specific benchmarks, with Qwen represented in light blue and GPT-4o in gold/brown. | Benchmark | Leader | |---|---| | GPQA Diamond | Qwen3.5-4B (Significant lead) | | MMLU-Pro | Qwen3.5-4B | | MATH-500 | Qwen3.5-4B (Largest lead, nearly 95%) | | MMMU-Pro | Qwen3.5-4B | | Video-MME | Qwen3.5-4B | | MMMLU | GPT-4o (Slight lead) | | MMLU | GPT-4o (Slight lead) | Key Takeaway The graphic highlights that the much smaller 4B parameter Qwen model from 2026 outperforms the older 2024 flagship GPT-4o in specialized reasoning and math tasks, while GPT-4o maintains a narrow edge in general knowledge benchmarks like MMLU and MMMLU. Would you like me to analyze the specific percentage gaps for any of these individual benchmarks?

at least on benchmarks, Qwen3.5 4B beats GPT-4o

GPTQ 4-bit quant means it fits into 2 GB

You can tell they never read nor studied anything by the people who lived through wwii or Vietnam

Rapidly rebranding all my search benchmarks as eval awareness benchmarks

I curse at it but only after I ask it to summarize the current relevant files and functions… then I start a new conversation

I think we’re screwed if they give the tools the ability to remember…

This is pretty amazing. Could flip the vast swatches of rural America to EVs

Imagine building a modest solar farm and some battery and capacitor banks… and the rural residents could indefinitely power their vehicles with a short stop, and never have to truck in gasoline— completely self sufficient

Does anyone know exactly how the new interrupt modes work on the latest models? Are they just interrupting the context and appending the interruption with special tags or something?

A line graph titled "GPT-5.4: 1M Context Reality Check" showing needle-in-a-haystack accuracy (MRCR v2, 8-needle) across different context window ranges. The accuracy starts at 97.3% for the 4-8K range and remains relatively high until 128-256K, where it begins a sharp decline. In the final two ranges, highlighted in red as the "1M context" zone, the accuracy drops significantly to 57.5% (labeled as a "40pt drop") at 256-512K and falls to 36.6% at the 512K-1M range. The source is cited as OpenAI GPT-5.4 eval table, dated March 5, 2026.

GPT-5.4 has 1M token context! wow!

reality:

Does local law clarify though that it’s being used to circumvent another nations rights protections? Seems like they would have had grounds to oppose this if they knew