Tron hauga (@Tronhauga) header
Tron hauga (@Tronhauga) profile picture

Which trilogy are you named after? Cofounder of @trynativeapp (YC F26)

San Francisco
Joined September 2025
Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)Photo by Tron hauga (@Tronhauga)
Three years ago, @tronhauga was my boss at a startup organization in Norway. That's where I first saw how startups actually work. Now I'm crashing with the guys at @trynativeapp in SF, building my robotics startup. Crazy how one person can shift your path.
2
1
3
77
Tron hauga retweeted
Composal (YC F26) sends a swarm of agents through your product to test it the way real users do. Your coding agents build it; ours break it, record the evidence, and hand your agent what it needs to fix and re-verify.
49
20
7
212
133,099
The best part about YC is dinner (no fat jokes please)
1
1
10
299
Does YC cap you on nr of launch videos allowed in one batch?? Native 3.0 coming in hot
3
2
17
565
Tron hauga retweeted
Frontier coding performance at up to 20x lower cost. Introducing the world’s most cost-efficient harness: Reason Stop burning thousands on runaway agent bills. Reason outperforms top harnesses on Terminal-Bench and DeepSWE while dropping cost per task to a fraction of the frontier baseline. Comment “Reason” and we’ll send you early access + free compute credits. Try for free: reasonmachines.com
1,192
493
223
2,102
3,364,116
ChatGPT and Canva just had a baby 💥 And today it was born inside @trynativeapp 🚀 The posts go out under your name, so you should be able to change them however you want. With Edits you change the words, photos, elements and layers. By hand or via the assistant 🔥 Let's go!
4
2
7
271
There's a 50% chance I f**ked up at OpenAI today I met this one researcher in post-training for accounting. He told me he never met anyone who automated their bookkeeping 100% before me, and that LLMs still do some obvious mistakes. I'm both flattered and worried.
2
1
15
307
Launching @reific (backed by @ycombinator): RL environments for superhuman reasoning over time. AI has learned to reason over text and code. But most of what matters in the real world shows up as numbers that change every second: a heartbeat, a bioreactor, a GPU cluster. No human can track thousands of signals at once, but AI should be able to. In our environments, data streams in on a clock while the agent works, and every action is graded by what actually happened. Most AI training today teaches models to match human experts. We treat human performance as the floor. Today, even the best frontier model averages under 50% on each of our three task families: personal agents, industry, and science & compute. Training an open model on our environments raised its full-solve rate on unseen tasks by 54% (from 25.6% to 39.5%), with no loss on general benchmarks. I’m a published time-series researcher who simulated superconductors at CERN and shipped AI to hyperscalers as co-founder and CTO of Entangl (YC S24). I’ve seen firsthand how models that ace text and code fall apart on live signals, so I started Reific to build the infrastructure that keeps AI climbing the frontier, long past human level in the real world. If you train agents, try to break one of our environments: reific.co
30
9
1
83
70,576
Tron hauga retweeted
Today we're launching Petrichor (YC F26), a wet lab for frontier AI labs. We run experiments in parallel and return the data from every step.
73
33
28
372
85,294
This is why Canva fired 1200 people Grok bot working inside our canvas. All our subs will get their own virtual machines, making always-on agentic marketing a reality. Works really well w Dots too.
1
5
170
You became a plumber to fix pipes. An electrician to fix the lights. A hairdresser to cut hair. A mechanic to fix cars. A dentist to fix teeth. A baker to bake. A dev to code. Not to do marketing. Go do what you're best at and we'll handle the rest. Link in comments 👇
3
2
10
484
We got into YC and now I feel like a loser. I net 900 steps a day. And work from 8AM to 2AM. But the product's getting better, churn's going down, and pilots are being signed. Got that dog in us. 30 days left, alarm set for 7. Vamos!
22
1
130
9,001
OpenAI is the most YC-pilled company in the world. We've spent over $100,000 on tokens with them. A couple of months back, their new image model saved our business. Thanks for backing founders early, Sam, Katie and Apoorv 🙏
5
2
16
573
F your AI b2b SaaS. We just launched Palantir for booking meetings. The last 14 days, companies like @MonacoGTM , @Nowadaysai, @Rippling , @salesforce , @marketercom, @posthog , and @retellai spent $100k+ with us on literal cakes. From today, Daymaker finds your prospects, reads their entire digital footprint, buys the one gift that makes their day, and gets you the meeting. Comment your name and I'll reply with everything Daymaker finds on you and the gift it would send you. If you dare. 🎁
355
68
114
902
336,855
Tron hauga retweeted
we are joining Founders Hub! excited to build alongside the top founders in Norway!!
6
3
24
561
Food distribution will be needed no matter how far AI advances. But the industry runs on a 2.9% profit margin, and I see that as a problem. The solution is efficiency. Apps and software have been tried before, but they never took hold in an industry where paper and texts are the norm and technology is not a strength. Asakana (YC F26) automates sales and the entire back office without changing how anyone works. Keep using email, text and handwritten paper. No ERP change needed. With minimal change in behavior, we make food distributors more efficient and more profitable. asakana.co/?utm_source=x&utm…
78
59
24
318
62,634
Today, we (@dreamscalelabs) are launching Dreamscale. The fastest cloud inference platform for robotics. Dreamscale lets you run VLAs and WAMs from the cloud with ultra-low latency. We optimise across the full-stack, from the kernels to network for real time control. We’re making shared endpoints generally available today on dreamscalelabs.com 1/N
9
2
2
28
3,419
Tron hauga retweeted
Every day this week, we're launching a new feature @GrepdotAI to help you automate the work that's slowing down your business. 𝗗𝗮𝘆 𝟮: 𝗗𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗼𝘂𝘁𝗽𝘂𝘁𝘀 𝗽𝗼𝘄𝗲𝗿𝗲𝗱 𝗯𝘆 𝗝𝗲𝘃 Customers often ask why our agents don't attach confidence levels to their recommendations. My answer has been that LLM confidence levels don't mean much. The model writes a number after the answer, predicting what a confident-sounding score looks like, with a bias toward making humans happy. RLHF pushes models overconfident. Decision models like Jev by @typesafeai change that. You declare the options up front. Jev returns a probability for each one in a single pass, plus a confidence score derived from how concentrated that distribution is. TypeSafe trained it with RLCD, which rewards calibration instead of rater approval, so decisions at 0.9 should be right about 90% of the time. In Grep, you can now build agents that use frontier models for the research and diligence, then hand their findings to Jev to make the final decision against your criteria. You set the threshold: above it the agent acts, below it a human reviews.
16
2
4
34
2,334
Today we're launching artificial intuition. A model that predicts which idea will stand out, before anything is published. Already live in Native for 2,200+ businesses. API waitlist is open: native.no/norn 🚀
65
17
6
217
11,806
Getting your inference right shouldn't be hard. @useRouterPlus (YC F26) is live: the right model and provider for your workload, at the best price. check out : routerplus.com/capacity
133
47
78
447
266,327