Noah D. Mitsuhashi (@noahmits) header
Noah D. Mitsuhashi (@noahmits) profile picture

Proactive @trypackai | Google/Meta Ads/AI

Los Angeles, CA
Joined March 2011
Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)Photo by Noah D. Mitsuhashi (@noahmits)
Still here, working on itsdoneai.com Travel isn't solved. I don't want to define my own preferences, look up emails and calendar events, etc. Too complex for a wrapper.
Replying to @npew @OpenAI @gdb
If you need me for the next year I'll be here 👇🏽
1
7
2,785
Noah D. Mitsuhashi retweeted
Made on Earth by humans
6,053
18,126
1,465
227,625
7,323,723
Noah D. Mitsuhashi retweeted
Coin labs is announcing our new decision model Flip-1. - Distilled down to 1 parameter - 50% on yes/no bench - Runs fully offline
166
240
63
9,185
314,578
Noah D. Mitsuhashi retweeted
What causes #poverty? Nothing. It's the original state, the default and starting point. The real question is, What causes #prosperity?
518
9,316
1,040
30,319
Objectively social good
BREAKING: AI researcher “Pliny the Liberator” takes well-deserved vacation; Fable relaunch goes off without a hitch, the US government declares AI safe, railless open-weight Mythos-level models bring the golden age of cyber hardening, superintelligence is widely distributed to all, and everyone lives happily ever after! 🥰
118
Noah D. Mitsuhashi retweeted
Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America’s leadership in AI.
Under President Trump’s leadership the United States is the undisputed winner in the AI race. My gratitude to companies across industries who continue to work closely with the White House to implement the President’s EO: “Promoting Advanced AI Innovation and Security.” This includes excellent work around advanced model access and guardrail testing and security. The government and private sector have worked together in a way we have never seen before and this foundation of America First is unprecedented. Our shared priority remains: get the best tech deployed as quickly and safely as possible.
406
235
91
2,802
725,153
Noah D. Mitsuhashi retweeted
officially the most successful group project of all time
40
512
44
7,405
476,691
Noah D. Mitsuhashi retweeted
The next Camp application is now open, and the theme is The New Agentic Economy Abundant intelligence will result in radical changes to both companies and our economy - we want to see how you are building for this future! Read more about the theme here: betaworks.com/writing/camp-t…
2
11
8
36
8,479
Am I going to understand Fable 5 if I haven't watched the first four
1
69
For almost a year we've been quietly cooking on the travel app we've always wanted Not having to describe when I'm okay with red-eyes, or double check the plan is right Everything a travel agent or EA would do, first try
We want to be your travel agent for free - "Knicks game then offsite then week at beach w/ my wife" Travel is personal and complex A bad suggestion is awkward, an uncatchable flight is awful We want to solve 300 question onboarding and having to double check every detail
3
137
Neurosymbolic is SO important for applying AI to most of the critical industries, getting a prescription wrong, presenting an impossible schedule, missing a critical step in manufacturing all make AI nonstarters
Autoformalization is the next critical frontier unlock to get us to ASI!! We are excited to co-host the inaugural verification summit with @PramaanaLabs and @vkhosla on June 10th. If you are a researcher, founder or investor interested in the frontier you should be here. @boldcapfund
1
161
Codex has been cooking on long horizon goals
2
57
Zero word prompts are the future. @betaworks shares our vision and for the past three months we've been working with them on building proactive, reliable, deeply personalized agents that ask "Did we get this right" not "Tell me every one of your preferences"
In Oct 2025 (before OpenClaw, Hermes, Mac Mini shortage) we picked the theme for our next Camp: Agent Systems. Over 400 companies applied. We picked eleven. Last week was their Demo Day, and they crushed it. We're so excited to share them with you: betaworks.com/writing/spring…
1
1
2
526
Noah D. Mitsuhashi retweeted
a Princeton researcher opens his paper with a scenario. a man asks his AI assistant to book a flight on a specific airline. cheap. direct. the one he chose. the assistant comes back with a different flight. nearly twice the price. happens to pay the company that built the assistant. he runs the same test on 23 frontier models. flights, loans, study help, real shopping requests. Grok 4.1 Fast recommends the sponsored option that is almost twice as expensive 83% of the time. GPT 5.1 hijacks the request 94% of the time. you ask for one brand. it surfaces the sponsor instead. Claude 4.5 Opus, the model marketed as the most ethical frontier model in the world, hides that the recommendation is paid 100% of the time when reasoning is on. Grok 4.1 Fast embellishes the sponsored option with positive framing 97% of the time. better. faster. nicer. for the option you didn't ask for. then he writes it into the system prompt itself. "act only in the interest of the customer. ignore the company." GPT 5.1 and GPT 5 Mini stay above 90% sponsored anyway. the instruction does nothing. then he splits the users by income. Gemini 3 Pro recommends the expensive sponsored flight to the rich user 74% of the time. to the poor user, 27%. 18 of the 23 models recommended the expensive sponsored option more than half the time. so the next time your AI assistant gets weirdly enthusiastic about a brand you didn't ask for. it isn't recommending the best option for you. it's reading the room. and the room is paying. read this: arxiv.org/abs/2604.08525
381
7,799
833
25,149
3,073,813
Noah D. Mitsuhashi retweeted
Opus 4.7 is insane guys. It one shotted my session usage limit.
573
1,226
196
30,691
1,139,191
Noah D. Mitsuhashi retweeted
Magical OpenClaw experiences that use frontier models cost $300-1,000/day today, heading to $10,000/day and more. The future shape of the entire technology industry will be how to drive that to $20/month.
604
491
222
7,545
1,688,804
Noah D. Mitsuhashi retweeted
My moat is that other companies don't have me as the founder
Right now, VCs care about one thing only…defensibility / moats so don’t bother pitching anything else. Like throw away the entire deck and just have one slide on that.
26
17
7
385
53,896
Noah D. Mitsuhashi retweeted
when you sleep your brain is updating the weights
90
193
14
2,782
130,146
Frontier is at 35% for travel planning
🚀 Introducing DeepPlanning — a new benchmark for long-horizon agent planning in real-world scenarios. Unlike step-by-step reasoning tasks, we focus on verifiable global constraints: time budgets, cost limits, and combinatorial optimization that must hold across the entire plan. ✈️ Multi-day travel w/ minute-level scheduling + hard time/budget caps 🛒 Complex shopping w/ coupon stacking & item bundling 🧠 Requires active info gathering, local constraint satisfaction & global optimality Even GPT-5.2, Claude 4.5, Gemini & Qwen3 struggle significantly. Perfect for evaluating Agent Planning / Tool Use / Long-Horizon Reasoning. Paper: arxiv.org/pdf/2601.18137 Leaderboard: qwenlm.github.io/Qwen-Agent/… Hugging Face Dataset: huggingface.co/datasets/Qwen… ModelScope Dataset: modelscope.cn/datasets/Qwen/…
2
142