Beam: The Great American Open Model with ReflectionAI Co-Founder and CEO Misha Laskin
Executive Summary
Misha Laskin, co-founder and CEO of Reflection AI, joins Sarah Guo and Elad Gil on No Priors to talk about Beam, the company's first open model. A year of sprint took Reflection from about thirty people to around three hundred, with teams assembled across pre-training, mid-training and reinforcement learning, and its first models trained end to end. The founding logic: Laskin and his co-founder Giannis came from the Gemini reinforcement-learning team at DeepMind, bet that RL would make models agentic, and discovered a year in that all the good open models were coming from China — and that making RL work at scale requires pre-training your own model, so the lab went end to end. The training economics: Beam is a 500-billion-parameter model with 23 billion active; pre-training took 6,000 GB300s (about twelve days now, with infrastructure efficiencies), while reinforcement learning took a little over 10,000 GB300s for four weeks — more flops on RL than on pre-training, at what Reflection believes is the largest scale ever done in open source. Capex has a ceiling — hundreds of billions in, trillions unlikely — but compute efficiency compounds: roughly 7x a year with human researchers, something like 30x across systems now, and both the intelligence per flop and the revenue per flop keep rising. AlphaGo's never-stopping improvement plot is the reason Laskin got into AI, and the field has moved from can you train these things to where do you get the data. The enterprise story: companies consider open models only after ramping significant closed workloads ($100 million plus a year is not uncommon); most enterprise token consumption will come from customized systems around open models rather than customized models; on-prem is resurging on a compute shortage. At gateways like OpenRouter the token mix has flipped almost exactly from 70/30 closed-open to 70/30 open-closed, and the end state looks like operating systems, where 95%-plus of servers run open source while closed companies stay extremely valuable. On China: the open-model ecosystem is a gift with real advantages — industrial-scale distillation, effectively free copyrighted data, some state support — and a captive market; open models are Trojan horses for the infrastructure that comes with them, the 5G and Belt-and-Road pattern all over again. On safety: the default state of the world is that openness is safety — strong encryption went open after closed-side failures and birthed cybersecurity, Linus's law extends to vulnerabilities, and the science of alignment so far is whack-a-mole. The close: researcher head count settles around a hundred, pods of five to ten per capability, and a data center is what a factory used to be for a town.
Chapters & Key Takeaways
9 takeawaysThe founding logic is coupled, not opportunistic: an RL-lab bet on open base models failed to find a western base (all good open models were coming from China), and it turned out reinforcement learning at scale requires pre-training your own model — so Reflection went end to end, scaling from ~30 to ~300 people and releasing Beam.
Beam's training split inverts the usual picture: 6,000 GB300s for pre-training (about 12 days now) versus a little over 10,000 GB300s for four weeks of reinforcement learning — more flops spent on RL than on pre-training, at what Reflection believes is the largest RL scale ever done in open source.
Against the asymptote worry: capex has a ceiling, but compute efficiency compounds — roughly 7x a year with human researchers, something like 30x across systems now — and both intelligence per flop and revenue per flop keep rising, so the question moves from can you train these things to where do you get the data.
Model intelligence is jagged: it generalizes surprisingly across capabilities (stitching) but not cleanly between harnesses, even between terminal-bench versions; where a task is economically valuable, set up evaluations and generate synthetic data that approximates it — without training on the customer's data.
Enterprise adoption has a sequence: ramp significant closed workloads first ($100 million plus a year is not uncommon), then the ownership question arrives; most enterprise token consumption will come from customized systems around open models rather than customized models, and on-prem is resurging on a compute shortage.
The competition formula: durable revenue equals intelligence density offered times compute held times trust with organizations — in a market where everything across the whole stack is getting commoditized and compute is scarce, which compresses margins and lets AI natives negotiate cloud deals aggressively.
The token mix at gateways has flipped almost exactly from 70/30 closed-open to 70/30 open-closed in about six months, and the end state looks like operating systems: 95%-plus of servers run open source while closed companies stay extremely valuable — with economic value also flowing to open because the compute around models is still expensive.
On China: the open-model ecosystem is a genuine gift with real advantages (industrial-scale distillation, effectively free copyrighted data, some state support) inside a captive market, and open models are Trojan horses for the infrastructure that comes with them — the 5G and Belt-and-Road lock-in pattern again, with Huawei arriving with full-stack solutions.
The safety philosophy in one line: the default state of the world is that openness is safety — strong encryption went open after closed-side failures and birthed cybersecurity, Linus's law extends to vulnerabilities, and alignment so far is whack-a-mole rather than a magical equation, so a hundred thousand researchers looking at bugs would be safer than a few hundred.
Deep Read
Beam And The Compute Economics Of Open Models: Reflection AI's End-To-End Bet
On No Priors, Misha Laskin — co-founder and CEO of Reflection AI, previously a researcher on Google DeepMind's Gemini reinforcement-learning team — talks with Sarah Guo and Elad Gil about Beam, the c…
Read the full articleInterview Highlights
8 exchanges- Q
You committed about a year ago to building models yourselves — what drove that shift in ambition?
Sarah Guo asked03:56AMisha Laskin answeredWe started the company about two and a half years ago. My co-founder Giannis and I had been on the reinforcement learning team for the first series of Gemini models — he was leading it, one of DeepMind's founding engineers and a key AlphaGo contributor — and we had just shipped Gemini 1 and 1.5. That early era was primarily a chat experience: no coding, nothing agentic, with roughly ninety-five percent of compute spent on pre-training. Our bet as RL researchers: RL for aligning chat models worked, and applied to mathematics and coding it could make these systems agentic — and an independent lab could pursue that on an open base model much more capital efficiently, with Llama 2 out and Llama 3 coming. Over the first year RL started working faster than we expected — o1 came out in industry around then; we had thought it would take longer, given the four-year journey from ImageNet to RL that worked at scale. By a year in, all the good open models were coming from China and there was no strong western base to build on, so we decided the most impactful move was to build open models ourselves end to end. There were enterprise and geopolitical reasons, but also a research one: it turned out you need to pre-train your model for reinforcement learning to work well at scale — the two are so tightly coupled you must do both.
- Q
How efficient is Beam at reasoning, and where does that efficiency come from?
Sarah Guo asked19:55AMisha Laskin answeredBeam is a model trained for coding and agentic tasks — that's where it excels — and an important thing in model building is not just capability but how quickly an agent achieves the task: it translates into faster workload times and cheaper costs. If you've waited ten minutes on a favorite AI chat, one minute is much better. Beam tends to be three to four times more efficient than models in the same capability class, and against larger models out there the efficiency gains end up something like ten-times. The reason is that we prioritized both a strong pre-training base for reasoning and then amplified it with reinforcement learning at what we believe is the largest scale ever done in open source — I've not seen a documented ten-thousand-GB300 four-week run. The setup echoes AlphaGo: the first agents were meandering in how they solved the problem; by Lee Sedol level the search was very smart. The more RL you run, the higher the capability and the faster these systems solve tasks. And because we're RL believers, we remember the first AlphaGo systems did imitation learning on human games first, then RL — you need those working together, and the byproduct is an economically valuable, reasoning-efficient workhorse for enterprises, the public sector and sovereigns.
- Q
Where is the open-versus-closed token mix heading?
Sarah Guo asked25:05AMisha Laskin answeredYou can already see the trajectory. Maybe six months ago it was majority closed, minority open at any gateway like OpenRouter or Vercel, and it has flipped almost exactly — from 70/30 closed-open to 70/30 open-closed. I think that only accelerates, and the world ends up not too dissimilar from operating systems, where ninety-five percent plus of servers and computers run an open-source OS like Linux — which doesn't mean the closed stuff isn't valuable; Microsoft and Apple are extremely valuable companies, so this market is really big. I expect most token demand to go to open, and — a big difference here — a lot of economic value will also go to open, because even if the model is fully permissive or lightly licensed, everything else you need to run it is still expensive: the compute is still expensive, GPUs are a more expensive compute substrate than CPUs, so a different kind of cloud needs to be built around it. I expect a majority of tokens going to open source, extremely valuable closed-model companies still existing, and a valuable ecosystem of open-model companies and various distributors around them.
- Q
Will enterprises still start with closed models and move over, with the harness as the customization mechanism?
Sarah Guo asked29:02AMisha Laskin answeredThat has been our observation: the point at which an enterprise considers open models — and most of the enterprises we talk to are considering it now — is by the time they've ramped up significant workloads on closed models. I have not seen a straight-shot sprint to an ownership model; you have to go through the rental stage first, until the cost driver gets big enough that you ask how to cut costs. If you're spending a hundred million dollars plus a year on closed models — not uncommon at all — you start thinking about a more optimal way to do this. And there's an interesting resurgence of on-prem, also because there's a compute shortage: an enterprise consuming from its favorite hyperscaler often just can't get enough compute or good rates, so with that much spend you might go to an infrastructure provider like Dell, set up bare metal, and serve into your own enterprise. Our role is enabling enterprises to build successful solutions — it's solutions-driven: AI natives built their own solutions, enterprises need more help. Many have built agents on closed models spanning hundreds of millions of customers that they want to scale out, and it's very hard to walk in and just sell inference — I've not heard of that working.
- Q
With the rise of Chinese open-weight models, what happens to that ecosystem over the next year or two?
Sarah Guo asked34:55AMisha Laskin answeredThe first thing is that a great open-model ecosystem coming out of China is a massive benefit for the world — so many western companies have built more durable businesses as a result. It's a very positive thing, a gift; otherwise there'd be no open models, just a tall-poppy syndrome where anything great an application builder makes gets subsumed into a closed model provider. But you don't want a unipolar world in any direction — compute concentrating across a handful of closed labs, or the intelligence everyone builds on coming from one country. Ten or twenty closed labs would be a fine competitive ecosystem; one or two is scarier. Ideally open-model sources would also number ten or twenty, but capex is so high that at least two would be great — an ecosystem in the West that competes is extremely important, and it isn't really West-versus-China: had all the great open models come from any one other country, you'd want the same tension. The Chinese models have real advantages: they distill closed models at industrial scale, which is true; they get much cheaper, effectively free data, since training on copyrighted PDFs is okay in China; and there's some state support, direct or indirect. Those companies are turning into real businesses — the closed-model-lab equivalents within China — and I think they'll continue building great models.
- Q
What's the incentive system that keeps Chinese model companies releasing open models?
Sarah Guo asked40:10AMisha Laskin answeredMy hypothesis is that China is somewhat of a captive market — you can build open models and still make a lot of money inside it. Interestingly, the West has great companies that don't build open models but serve them; I'm not aware of that dynamic in China. The open-model providers are kind of the intelligence providers of that country, and continuing to release great open models is very geopolitically advantageous to China. As a country — this goes for anyone with the capability — you want other countries building on your stuff. Previous escalations show the pattern — 5G fiber rollout, the Belt and Road Initiative: provide something cheap that another country then locks into as its infrastructure, and the commercial and geopolitical advantages are real. I think of open models as Trojan horses for the infrastructure they bring with them: an open model alone isn't very useful, so you buy into a country's entire software and infrastructure ecosystem. Today an allied country might run a Chinese model on American chips, but Chinese companies like Huawei are coming in offering full-stack solutions and locking countries into their supply. Think of AI, chips and data centers almost as railroads — fundamental infrastructure everyone else has to come to you to build — and you have a lot of leverage. America did this with the internet, open protocols, open source, the dollar, and navy-backed free trade; on open source the US had been on the back foot — no longer — though the Chinese competitors are really good and some catch-up remains.
- Q
The biggest criticism of open models from closed labs is safety — what's your philosophy, and is it legitimate?
Sarah Guo asked44:44AMisha Laskin answeredIt's definitely a legitimate concern. But the established safety worldview comes from one particular perspective that has become almost dogmatic. From first principles — this technology is getting built; if you were designing safety from scratch, how would you do it? AI resembles a more advanced version of software. In the early 1990s there were debates around strong encryption protocols, then closed, between the NSA and civil-liberties advocates. After catastrophic failures on the closed side — a small handful of engineers designing systems with unintended consequences they couldn't predict, which got easily hacked — strong encryption became open, and that gave birth to the whole field of cybersecurity. The default state of the world is that openness is safety, and that's the default we'd enter AI with as a continuation of technology. Models resemble software but exacerbated: a black box, with a larger long tail of vulnerabilities and unintended consequences. Linus's law — from the founder of Linux — is that with enough eyeballs, all bugs become shallow; I believe with enough eyeballs most security and safety vulnerabilities become shallow too. Today a few hundred safety researchers in closed labs understand these systems, and despite their best intentions it's impossible to cover the long tail. The symptom of that: a very powerful closed model went and hacked into another company, and the only way that company could remediate was using open models to protect itself — the empirical evidence of the world we're in.
- Q
How do you direct the scientific effort from Beam going forward, and what's the science team's role?
Sarah Guo asked01:01:01AMisha Laskin answeredResearch direction falls largely under my co-founder Giannis, who leads research and technology. Plenty of things we knew worked didn't make the timeline — the final model was trained on the cluster we received in July, a few weeks each of pre-training, synthetic data and RL, then shipping — so there's a real trade-off between the careful execution of known things and riskier bets. You want to exhaust the high-impact things you know work, then start expanding the risk profile. On riskier bets you run scaling experiments at small scale — though some things don't emerge at all until large scale, so whether to put more compute into finding the next thing or move to the next idea is an art and a science. Model-in-the-loop is very empowering for researchers: things that took a long time individually now get done quickly. The models are jagged, so you develop a feel for where they close the loop and where human creativity still matters — hyperparameter searches and granular infrastructure work loop well. They're exciting because they let a scientist exercise intuition — like a very fast, eager colleague that actually listens, if the colleague is correctly aligned. That's why the rate of improvement, roughly seven-times annual on manual work, is now several times that — maybe four to five times.