
entrepreneur, now vc | founder @kindredventures | early investor š± @uber @coinbase @perplexity_ai @fal @get_hydrahost @meckaAI @parasail_io @nuro @aaloatomics
san francisco, ca
Joined February 2008
- Tweets19.8K
- Following4.5K
- Followers42.6K
- Likes40.8K
Pinned Tweet
If you hit a wrong note, it's the next note that you play that determines if its's good or bad. -Miles Davis šŗ
steve jang retweeted
Boreal-H3 is now #3 on the @ArtificialAnlys AA-Video-I2V v2.1 leaderboard (with audio), at roughly 1/14 the cost of the #1 model.
The ranking comes from people voting blind on clips from 36 models.
That price gap matters for ad teams. Good ads come from testing many ideas, and every extra take has a cost. At our limited-time price of $0.04 per video-second at 1088p, a 10-second clip costs $0.40, so a team can try ten versions of a concept for $4.
Owning the model also lets us train it on what our customers run into, like a product label that warps mid-shot or a face that shifts between scenes.
Leaderboard: artificialanalysis.ai/video/ā¦
Try it: app.creatify.ai/tool/ai-toolā¦
multi-silicon inference š
š
š
New silicon coming to the Parasail fleet: @positron_ai Titan, which offers 4X better cost-per token than GPUs.
If you build on Parasail, nothing changes in your code. Same API, same Elastic Endpoints. What changes is underneath: another chip in the fleet, and more ways to route each workload to the hardware that runs it best.
steve jang retweeted
Positron's Titan is joining the Parasail fleet. 5MW. Trillions of tokens a day of cost-optimized compute by late 2027. @parasail_io @positron_ai
@parasail_io is the first multi-silicon inference cloud that routes your prompts to the best chip for the job, serving trillions of tokens a day to hundreds of customers. Positron fills a vital part of compute spectrum on the cost optimized side: 4X better cost-per token than next gen NVIDIA/AMD systems which is perfect for large-scale agent swarms, background agents, and physical AI.
Positron has done incredible work building a chip and system that orchestrates across DRAM and SRAM allowing for massively scalable memory pools with commodity memory giving the best of both worlds for memory capacity and bandwidth.
We've been excited to partner with them in the early days through the launch of their Atlas chip and now big capacity plans with Titan.
@positron_ai
the š§µ below is fascinating preview of whatās to come in autonomous systems design using AI - in this case, custom silicon.
@architectlabs has deployed agent swarms to co-design a higher performance CPU from scratch. Their AI agents took an existing model of computation (interaction nets) and co-designed a compiler, a new instruction set, and the processor itself, through verification and a final 7 nm chip layout.
theyāll release the RTL and ISA, and provide a FPGA-hosted playground for devs to run programs on.
soon, using AI systems like Architect, youāll be able to design and own your own chips, customized for your workloads. āØ
Introducing Frostbite: a new class of CPU built around interaction nets.
Given only interaction net theory, our agents co-designed the ISA and chip architecture from scratch, running large swarms to design, verify, and optimize the CPU down to the final GDS.
On our first set of target workloads, Frostbite is 4.6x more performant per watt than modern traditional CPUs at similar area.
šš½šš½šš½
Proud cousin! @km giving the keynote speech to open the āSouth Africa and the AI Opportunity Conferenceā here in Silicon Valley. Born in South Africa and left the country as a child in exile as a refugee in the 80s, excelled at school, studied at Stanford, and then - in his own words ābankrupted his first few companiesā - before going to to become a hugely successful venture capital investor. That his story is exceptional but far from unique, is exactly what this place is about. This is the story of Silicon Valley.
1
12
1,236
steve jang retweeted
Introducing Frostbite: a new class of CPU built around interaction nets.
Given only interaction net theory, our agents co-designed the ISA and chip architecture from scratch, running large swarms to design, verify, and optimize the CPU down to the final GDS.
On our first set of target workloads, Frostbite is 4.6x more performant per watt than modern traditional CPUs at similar area.
steve jang retweeted
Yesterday, I revealed that Perplexity crossed $1B ARR. We have strong margins and a healthy growth rate. We donāt talk about our finances very often because weāre focused on building. Follow our research: research.perplexity.ai
Huge congrats team @kotoba_tech and @ElevenLabs! š¤š½š¤š½š¤š½
weāre honored to have backed you Nori and Jungo, and doing maybe the most fun live demos ever. ;)
do great things together in voice AI and agents in all of AsiaPac, @mati @noriyuki_kojima @jungokasai. é å¼µć!!!
Today, Iām incredibly excited to announce that Kotoba (@kotoba_tech) is joining ElevenLabs (@ElevenLabs) to push the boundaries of voice AI in Japan and East Asia.
From day one, Kotobaās mission has been to build world-class AI models and products from Japan that make a meaningful difference in peopleās lives. Iām incredibly proud of what weāve built over the past three and a half yearsāfrom our work on Agentforce and voice AI models powering products at major global companies to our collaboration with YOASOBI and the popular Kotoba translation app. In that time, we grew from two people to a team of 15 and raised more than $23 million, all while staying lean and focused on bringing our research into peopleās hands.
As proud as I am of what weāve achieved, Iām even more excited about what we can accomplish with ElevenLabs. With Kotobaās research talent and deep understanding of the region alongside ElevenLabsā global leadership in voice AI, weāre building a true dream team!
Voice AI is opening up new ways for people to interact with AI and connect across languages. People and businesses in Japan and East Asia deserve the very best of itāand weāre committed to delivering it.
Together, weāre taking a major step toward building a frontier AI lab in Tokyo: developing world-class voice AI models for a global audience and accelerating localization to bring the best voice AI products to Japan and East Asia.
A huge thank you to Mati (@mati) and Piotr(@dabkowski_piotr) for envisioning this partnership with us.
And to the Kotoba team, our investors, partners, friends, and familiesāthank you for believing in us and supporting us over the past three and a half years.
Chapter 2 starts now!
Nori
First off, we're so stoked to share that @KindredVentures joined Noeri's $110 million round, led by @MenloVentures, investing alongside @MadronaVentures, @WingVC, and @NEA, along with strategic investors @NVIDIA, @Snowflake, and @Cisco.
We had already been a fan of @YejinChoinka's and @guestrin's storied AI research work, but the first meeting with Carlos and lunch with Yejin went something like this:
Imagine a creative AI system that...
learns on the job
builds its own tools
trains its own models
observes and tracks progress towards the goals
you own and steer it to your specialization and needs
Do you believe people and companies need this especially in light of the few big (but not always better!) options out there in today's AI model and tool landscape?
So today, enter @noeri_ai and its brilliant founders @YejinChoinka and @guestrin. Most of the planet has used and loved products influenced directly by their collective research at @Stanford, @UW, @nvidia, and @Apple. We were so drawn to the combination Yejin and Carlos bring to Noeri: advancing the science of how AI continuously learns and building the tools and production systems that put that research to work at scale.
At @KindredVentures, we believe that human work should continuously elevate and refine the intelligence supporting it - and vice versa. Carlos and Yejin's vision for specialization and continuous learning, in an environment of ownership and creativity, holds immense untapped potential for a future of a plurality of intelligences. Stoked to be a supporter and backer of this grand mission!
Through my wonderful time at NVIDIA and Stanford, one question has kept me up at night:
Should the future of AI be monolithic or plural?
To answer this question, weāve raised $110M to get started, backed by @MenloVentures, @MadronaVentures, @NVIDIA, @NEA, @KindredVentures,@Wing_VC, @aixventureshq, @Principal_VC, @Cisco, @Snowflake, among others.
Today, Carlos (@guestrin) and I are stoked to launch Noeri, a neolab to push the frontier of plural intelligence.
Noeri means āthe plural of intelligenceā or many intelligences.
The future of AI we root for is plural: a healthy ecosystem of many intelligences, in all sizes, strengths, and quirks.
Discovery, in particular, has always been plural: many curious minds, each devoted to their own question, often on roads less traveled.
Our path there is self-improving AI that:
-> learns continually on the job, closing the gap between what AI can do off the shelf and what the dynamic world demands
-> learns to use other AIs as tools and make new tools to overcome its own limitations
-> picks the right-sized model for your problem, at a fraction of the frontier cost
Then the power of intelligence, from the craftsmanship of making AI to ownership, can be in the hands of all.
@noeri_ai mission reflects who we are as the founding team. My career has been anything but linear, literally continual learning on the job. Carlos has devoted his career to building tools that empower people through AI.
We are all a bit contrarian, a bit adventurous, and above all, big believers in pluralism.
Join us to build this future together: noeri.ai/
a more detailed post by @TheYashKishore and I about their research and the mission:
kindredventures.com/announceā¦
2
337
steve jang retweeted
Unbelievably excited to back these brilliant, profound thinkers. Creators of XGBoost, nucleus sampling, and many of the most important pillars of machine learning.
They believe intelligence will take the form of many minds, owned by those who use them. We agree. My favorite expression of what AGI looks like is "spring in the garden"... Many flowers bloom.
So thrilled we get to back @noeri_ai. Congratulations @YejinChoinka and @guestrin. Onwards!!!!
Through my wonderful time at NVIDIA and Stanford, one question has kept me up at night:
Should the future of AI be monolithic or plural?
To answer this question, weāve raised $110M to get started, backed by @MenloVentures, @MadronaVentures, @NVIDIA, @NEA, @KindredVentures,@Wing_VC, @aixventureshq, @Principal_VC, @Cisco, @Snowflake, among others.
Today, Carlos (@guestrin) and I are stoked to launch Noeri, a neolab to push the frontier of plural intelligence.
Noeri means āthe plural of intelligenceā or many intelligences.
The future of AI we root for is plural: a healthy ecosystem of many intelligences, in all sizes, strengths, and quirks.
Discovery, in particular, has always been plural: many curious minds, each devoted to their own question, often on roads less traveled.
Our path there is self-improving AI that:
-> learns continually on the job, closing the gap between what AI can do off the shelf and what the dynamic world demands
-> learns to use other AIs as tools and make new tools to overcome its own limitations
-> picks the right-sized model for your problem, at a fraction of the frontier cost
Then the power of intelligence, from the craftsmanship of making AI to ownership, can be in the hands of all.
@noeri_ai mission reflects who we are as the founding team. My career has been anything but linear, literally continual learning on the job. Carlos has devoted his career to building tools that empower people through AI.
We are all a bit contrarian, a bit adventurous, and above all, big believers in pluralism.
Join us to build this future together: noeri.ai/
steve jang retweeted
Meet Creatify Max, the newest member of your ad team
Everyone gets their own Max. It watches your accounts 24/7, writes the briefs, makes the videos, and launches the campaigns
Max does the work. You make the call.
more from bloomberg: bloomberg.com/news/articles/ā¦
317
What stands between autonomous vehicles and mass adoption? According to @KindredVentures's @stevejang: Manufacturing capacity, charging infrastructure, regulatory complexity, and the ability to deliver a great user experience at the right price. Listen to Ep3 of Unsupervised with @dfergusonnz for more: unsupervised.buzzsprout.com/
#drivenbynuro #selfdriving #autonomousvehicles #unsupervisedpodcast
steve jang retweeted
Introducing NovaAtom-Lite, an all-atom structure prediction model for drug discovery.
It predicts the structures of proteins, nucleic acids, and the molecules they bind, including how they fit together.
NovaAtom-Lite outperforms both IsoDDE and AlphaFold 3 on FoldBench antibodyāantigen structure prediction (DockQ ā„ 0.23):
NovaAtom-Lite: 79.07% (n=172)
IsoDDE: 75.58%
AlphaFold 3: 47.90% (n=167)
Our evaluation uses 5 seeds Ć 5 samples, with the top-1 prediction selected by the model's own ranking score.
Better structure prediction. A stronger starting point for your next experiment.
Live today via Geodesic API and NovaDDE.
š API: platform.geodesiclab.com/
š NovaDDE: agent.geodesiclab.com
š Model Card: geodesiclab.com/news/introduā¦
š @parasail_io rating top left in the highest tier! š
steve jang retweeted
Introducing teammates.
Teammates work for you. They take on jobs, and work proactively using your Computer tasks, skills, memory, and tools. Teammates collaborate with other teammates and can message coworkers in Slack.
Available now for Max subscribers on the web.
steve jang retweeted
Today we are rolling out the latest form factor for work: Perplexity teammates. Agents have reached the point where they can start to be trusted with responsibilities at work. Teammates excel at automating tasks like deep research, data analysis, QA, triaging tickets, and reviewing code, which has saved each of us internally at Perplexity hundreds of hours of menial work. Leaving us the much more interesting work of setting vision, judging quality, and teaching restraint.
Agents are now rising up beyond tools and task-based sessions to take on supporting and coworking roles. Teammates proactively reach out to you, other coworkers, and other teammates in the workspace. Teammates close the loop on their own work, finishing the job instead of requiring repeated steering or manual management of sessions.
Available to Max users as a research preview today.
fal doubled its valuation from $8B to $16B in just 7 months.
CTO @gorkem says the reason is simple:
"There is unlimited demand for the next productive token."
āIt's really that simple. People are spending so many tokens because they get value out of it."
"Advertising, marketing, e-commerce, retail, product photography. Hollywood is now actually spending tokens and money on this technology."
@fal
Today, weāre announcing we raised our Series E & F rounds for a combined total of $1.65B. Our Series E was led by @sequoia and our Series F was led by @BessemerVP with continued investment from @nvidia.
Weāre excited to push the boundaries of whatās possible even more next year and this funding will enable the team to build the way we want to.
From smartphones to self-driving cars to humanoid robots, Nuroās @dfergusonnz and @KindredVentures's @stevejang discuss the expanse of physical AI. Listen to Ep3 of Unsupervised wherever you get your podcasts: buzzsprout.com/2654995/episoā¦
#drivenbynuro #selfdriving #autonomousvehicles #unsupervisedpodcast





















