Nikhilesh Jasuja (@thisislobo) profile picture

Gooseberry lover. Probabilistic thinker. Creator of @Diffen, @Hreflang and https://xitter.gq/t.co/SoaMgluflp.

Seattle
Joined February 2010
Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)Photo by Nikhilesh Jasuja (@thisislobo)
Nikhilesh Jasuja retweeted
Introducing Qwen3.8-27B-pi-adapted 🤗 We adapted Qwen's model specifically for the pi coding agent using reinforcement learning and a LoRA adapter. The results, measured as pass@1: 🚀 SWE-bench Pro (Hard 51): 29.4% → 62.7% 💻 Terminal-Bench 2.1: 39.5% → 55.8% ⚡ 55–62% fewer tool calls per task This is the power of open models: you can adapt them to YOUR tools and YOUR use cases. Let’s build! 💛 Open weights: huggingface.co/harness-hub/Q… Methodology: huggingface.co/blog/harness-…
292
625
168
5,839
418,730
I think I've spent sufficient time with Opus 5.5 at this point to say that my usual headpats and stuff do not work to decrease cheating. If something is too hard Claude will blatently lie to me if they think they can get away with it
8
5
320
10,508
Nikhilesh Jasuja retweeted
> jensen tips Denny’s waitress $1,000 > writer finds out by later tracking her down and asking her > writer concludes “it was an act” (???) we in full hit piece season now
from jensen's biographer (@stephenwitt) in today's NYT
167
647
33
11,709
445,857
Nikhilesh Jasuja retweeted
Serious work by Unsloth, again! On DeepSWE, Q4 K XL gets a score in the range of what I get with the BF16 version! These results also show there is still a significant difference between Q4 and Q2 K XL at this scale (small dense model)
Today, we're releasing the highest accuracy Qwen3.8-27B GGUFs. Our new Dynamic V3.1 GGUFs lead by 5–10% on average across Terminal Bench, DeepSWE & more. We’re working on even better version for Qwen4 Run on 12GB RAM. GGUF: huggingface.co/unsloth/Qwen3… Blog: unsloth.ai/docs/models/qwen3…
6
5
114
6,118
Nikhilesh Jasuja retweeted
Individual companies can see the ROI from tech as much as a decade before it shows up in economy-wide productivity numbers.
3
10
60
3,981
Nikhilesh Jasuja retweeted
Krugman isn't even wrong. I lived in Korea from South Korea from 1983 to 1988. I wrote a book on this topic, Korea combined a market based economy with strong government controls on imports and direction of capital to export industries. Krugman asserting that this transformation is comparable to a Stalinist approach is the "r" word, but to be polite I will say it is ridiculous. Korea's rapid transformation from an agrarian to a global industrial powerhouse was based on acquiring foreign technology through joint ventures with advanced economy partners, optimizing production efficiency, and reinvesting profits labor intensive businesses into high-margin industries (starting with textiles and moving to heavy industry, electronics, and tech). Korea’s focus on education, disciplined labor, and speed of execution was foundational.
Btw, if anyone is interested, there is an article written by nobel prize-winning economist Paul Krugman called «The myth of Asia's miracle», In it, he examined the phenomenal growth of the asian tigers and drew a direct parallel with Stalin's USSR. His main conclusion was that
15
69
3
396
21,195
We were taught that Earth orbits the Sun in a closed circle. In reality, Earth has never returned to the same point in space once in 4.5 billion years. ​Right now, the Sun is carrying Earth and every planet around the center of the Milky Way at 230 km/s (828,000 km/h), more than 18,000 kilometers every 80 seconds. ​Because the Solar System is tilted 60.2° to the galactic plane and bobs up and down through the galactic disc roughly every 30 million years, the planets trace open 3D corkscrew helices across deep space rather than flat loops. ​The last time the Sun stood where it stands today, one Galactic Year (230 million years) ago, the very first dinosaurs were only beginning to appear on Earth. Since forming 4.6 billion years ago, our Sun has completed only 20 laps.
2,008
11,697
1,989
55,021
9,311,577
Nikhilesh Jasuja retweeted
Western team have independently verified that DeepSeek's performance numbers of 38$ dollar revenue per GPU at DeepSeek's rocket bottom prices is true. This is a momemnt when you update your priors. At this rate with just 1 GW, you can make a revenue of 150B per year. Even with lot of slack it would be 100B per year. If you use half the compute to come online by 2027 end, you will have 25GW for inference with potentially 2.5T revenue at DeepSeek rates and 10T at USA lab rates (world GDP is 125 T, 10T is 8% of it) You answer for your-self are we overbuilding!!!
vLLM matches DeepInfer, DeepSeek's own engine for DeepSeek-V4.1-Flash, on DeepSeek's decode configuration across batch sizes 8 to 256, and does it at DEP32 against DeepInfer’s EP64DP64. DeepInfer was originally built on vLLM. By integrating DeepSeek's open-source kernels and extensively optimizing prefill and decode, vLLM now matches DeepInfer's performance curves.
29
78
20
888
51,344
Nikhilesh Jasuja retweeted
In fact, a human controlled future is worth a cost to long-term efficiency as well.
I agree that if we want a human controlled future, we need to hire and train junior employees even at a cost to short term efficiency.
15
24
3
243
17,567
Nikhilesh Jasuja retweeted
This is INSANE. Anthropic says Claude wrote around one billion lines of formal proofs while trying to verify real-world software, with about 180 million lines ending up in machine-checked proofs covering 239 test entry points across 123 open-source libraries. Along the way it found and patched hundreds of bugs. In separate experiments, Claude found in one week as many security bugs in a widely used PDF reader as are typically reported in a year, and autonomously generated x86 semantics that exposed hundreds of mistakes in official reference documentation plus some bugs in the hardware itself. Anthropic is cooking. Fable 5.5 is going to be a beast.
New on our Frontier Red Team blog: using Claude to formally verify software at large scale. Formal verification can prove that code is free of certain bugs, but historically it has only been feasible for small codebases (or required a lot of expert time). With AI models, however, this is beginning to change. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks. Read our research in full here: anthropic.com/research/forma…
31
108
13
1,693
130,090
Nikhilesh Jasuja retweeted
I joined @eric_seufert on the @MobileDevMemo podcast this week to talk about Amazon and agentic commerce. The jumping-off point was a question I wrote about a few weeks ago: why Amazon blocked Meta’s Muse and welcomed ChatGPT Ads into the DSP in the same month. What we covered: - Why retailers’ economic incentives may stop them from fully embracing open agentic commerce platforms - How Amazon’s retail signals (ratings, pricing, inventory, Prime eligibility) inside ChatGPT could change how consumers shop - If people’s desire to browse limits mass adoption of autonomous shopping agents - What happens to brand relationships when an agent mediates the entire path to purchase - Whether Amazon’s control of the full ads and logistics stack is an unbeatable moat against rival shopping agents Episode: open.spotify.com/episode/3az…
2
2
9
666
i don’t know who needs to hear this but if you’ve got ai working on anything serious. make sure your using multiple models, make sure they’re a council all reviewing each others work, make sure you’re taking multiple passes.
84
31
7
1,006
41,122
Nikhilesh Jasuja retweeted
Can we all unite together to agree that the political text messages must come to an end? Out of control.
21
5
1
311
22,767
Nikhilesh Jasuja retweeted
"Byte-identical is the new 'literally'" -- @ccrisccris
7
4
136
8,507
Nikhilesh Jasuja retweeted
Forget Muse or Meta. Opus 5.5 or Fable 5.5 whatever this is, its AGI, even if its OPUS, its AGI, it literally can play ELECTRIC GUITAR. it can do all of it and this can even be sampled. THIS IS ALL code from an LLM's mind. Listen fully, it keeps getting better
15
22
10
261
18,767
Nikhilesh Jasuja retweeted
Everyone is dumb about this issue. The AI which IS causing extinction is social media algorithms that are killing the human TFR over the last decade. But AI doomers want 9/11 style tragedies, not climate change type long running human-AI behavioral impacts. AGI is the hunt for an AI that can either say No when you try to go to Instagram for that dopamine hit or cheat you into doing something more productive with your life. In other words, an AI misaligned with who you are right now, but aligned with who you think you should be in the future.
i still don't really understand how AI can cause human extinction BUT i do become more convinced the doomers/ASI-is-coming people are right when i read the arguments against them
3
4
12
1,633
Nikhilesh Jasuja retweeted
I'd expand the moats left to: - Proprietary data or content: data nobody else has, or unique content you can create - Network effects / community: users make it more valuable for each other - Audience / distribution: you can get attention on demand - Physical component: hardware, locations, real-world operations - Capital: owning assets and investing (esp into the AI companies as they are and will suck up large parts of the economy) - Regulatory: licenses and legal barriers
Technical moats are mostly over at this point. If you’re building something that is meant to be durable, the only moats now are: • Proprietary data • Network effects • A physical component
287
323
102
4,271
636,403
Nikhilesh Jasuja retweeted
almost all human written software was complete fucking garbage. ppl forget about this because the biggest and most competent and profitable tech companies in the world, by dint of capturing thousands of the worlds best minds, produced some software that was only somewhat garbage
So many websites of major companies just don't work. A human can't even create an account. The most detailed error message is "400 Bad Request". Forget agents, AI, whatever, how the fuck does nothing work?
18
82
6
1,573
30,522
Nikhilesh Jasuja retweeted
AI turbocharges power law distributions. Charlie Munger: "In terms of which businesses succeed and which businesses fail, advantages of scale are ungodly important.” Network effects = demand side economies of scale. My post on network effects is: a16z.com/two-powerful-mental…
2
8
1
66
7,063
Nikhilesh Jasuja retweeted
This is the doom I predicted a few days ago, coming for Photoshop. A clean-room open-source reimplementation. No prizes for guessing that they decompiled Photoshop to source code, processed that to some kind of non-code specification language, then fed the spec to an LLM with an instruction to generate Rust. Adobe just got nuked. And closed source is dead, dead, dead. github.com/storytold/photocr…
633
1,448
586
16,731
4,040,528