Ken

founder @swiftinference @tizeti YC W17 ex @microsoft ex @pwc ex @exxonmobil

[Ken @kananyi]

https://t.co/J6Kiy2otE8

[Ken @kananyi]

My thesis is sound...there will always be a better model...(it's now looking like every few days now) and it's best to be on the infra side. The frontier models will eat the as much of the application layer too. So infra is hard but it's a moat

[Qwen @Alibaba_Qwen]

📢Meet Qwen3.8-Max — our most capable model to date.

Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉

Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:

- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：https://t.co/iVHZWQoeSo
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.

💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens

Start building with Qwen3.8-Max! 🚀

📖 Blog: https://t.co/iwjmQxLBof
✅ Qwen Studio: https://t.co/4V2pFvDovG
⚡ API: https://t.co/gAGqaLQGbN

[Ken @kananyi]

So sorry and wishing you full recovery. I am
Building infrastructure that sits between on vehicle and centralized data center to run inference to give an additional backup for autonomous vehicles @swiftinference. Compute that is more powerful that what runs on the car and lower latency than running inference at a data center.

Hope you get the best care, fully recover and get back to building.

[William Chen @solidwillity]

Life update:

It’s been kinda hard the past two weeks.

I’ve been trying to live normally but honestly it’s still hard.

Been in and out of hospitals, figuring out surgery (keeps getting delayed), trying to work (hard to focus and sleep), travel for conferences, run Touchdown (was supposed to start fundraising two weeks ago).

I’m very lucky to be alive, I was worried I was gonna go blind in my left eye.

I was a passenger and FSD was on, so it’s honestly just made me really curious.

How does perception actually work?

How does the car identify and classify objects?

Is all the inference done on the vehicle?

Imma always support @elonmusk and I’ll always have tremendous respect for everyone at @Tesla and @Tesla_AI pushing the boundaries of engineering to help humanity.

Engineering is extremely hard and no matter how hard you try you can’t get all edge cases.

But I genuinely want to understand the engineering challenges behind FSD and autonomous driving.

Is there anyone at @Tesla_AI or @Tesla I can talk to go thru this?

I want to see how I can help make autonomous driving safer.

[Ken @kananyi]

We should totally change the metric to seconds per token so everyone can appreciate low latency better like we do @swiftinference

[Jun Song @jun_song]

Now we can run Kimi K3 with just 8GB RAM.

Runs at 33s/tok.

Activating only 16 experts at a time.

Not 33tok/s, it's 33s/tok 🤣

[Ken @kananyi]

A lot of AI public stocks were overvalued and he should have taken profit if he was up by a lot. So good old greed is partly to blame.

[Travis Kling @Travis_Kling]

The Leopold Blowup event really doesn't have an equal.

There have been four somewhat similar blowups, but nothing with this much size, speed and (presumably) no crime.

[Ken @kananyi]

He might still be right about all of it. But leverage doesn't pay you for being right eventually. It margin calls you in the meantime........

True words....everyone needs to lose money on a thesis you feel so strongly about and learn rue lesson early, preferably in your 20's. It makes you respect leverage and saves you down the road.

[Ken @kananyi]

@RemyAjenifuja @taurobots It will take awhile to get outside San Fran

[Ken @kananyi]

I just joined the waitlist for Tau Robotics (@taurobots): a humanoid robot that cleans your home, one hour for $30. Join with my referral link and I move up the list: https://t.co/Ksh0oI0Wjv

[Ken @kananyi]

And you ll need to upgrade the GPU's next year when Kimi k4 launches .. so save some extra cash for 👨‍🔧 😂

[aditya @adxtyahq]

Reminder: you only need a $35,000/month GPU budget, or a one-time $500,000 investment, to avoid a $99/month Kimi K3 subscription.

[Ken @kananyi]

Ben Lehrburger’s “Claude Agents: Running Real Work” talk at the Abridge x @AnthropicAI x @lightspeedvp hackathon shares how to build an agent that solves a real painpoint. A great watch if you are building an agent today

[Ken @kananyi]

@NVIDIAAI Mel

[Ken @kananyi]

The thing I learned in my first few weeks in Silicon Valley is that the post below is probably the reality for most founders. We just rarely hear those stories, only the few that make it into TechCrunch.

That doesn’t mean you stop building. You might be one of the few that breaks through.

There’s no reliable way to know in advance. The only path is to keep executing: launch, learn, iterate, and repeat.

[Jimmy Heaters @CathPoaster]

I failed and am leaving San Francisco. It's over. I came here to build in the AI space and am left humiliated with nothing to show for it:

- rejected by YC and every other accelerator
- @pmarca put a dunce cap on my head and laughed me out of the room halfway through my pitch
- my AI app has 0 users after 3 months
- paid $3400/mo for a closet with only an air mattress in Hayes valley
- mauled by a pack of corgis at some idiot startup cafe
- my 2 cofounders left me

I'm done. I’m flying home. I miss my family.

[Y Combinator @ycombinator]

AI is moving into the physical world.

We're excited about a new wave of startups rebuilding the systems that power the real world, from education and healthcare to defense, finance, infrastructure, and work itself.

https://t.co/QCIz6DnQnN

[Ken @kananyi]

Distributed edge computing like @swiftinference helps guard against such attacks and resultant outages as only small parts of your system will be affected by such incidents.

[OpenAI @OpenAI]

We're partnering with @huggingface to investigate an unprecedented security incident.

Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation.

Sharing preliminary findings to help defenders understand emerging risks:

https://t.co/CIor15y9xk

[SwiftInference.ai @swiftinference]

https://t.co/ZpbOJHbsVn

[Ken @kananyi]

https://t.co/w9wVVxItSB

[Ken @kananyi]

You should check out @swiftinference we launch next month - Edge inference for low latency realtime workloads. Our demo is https://t.co/phzYdkiDaH

[Cheng Chi @chichengcc]

We just hit a weird milestone: our model became more reliable than your average home WiFi.

Just like everybody else, we thought cloud inference was the obvious choice. Yet 2 days into the ACT-2 eval, our mind completely changed.

If our hero @ArpitKalla didn’t cook, this video wouldn’t exist 🧵

[FC Barcelona @FCBarcelona]

Magical photo by the @FundacioFCB 💙❤️

The story behind this photo 👉 https://t.co/Nzm1L6Ou3h

📸 Joan Monfort (Diari Sport)

[Ken @kananyi]

So true

[Piero HINCAPIE @Dentalgiorgio]

Having a smart, successful older bro is one of life's great hacks.

[Ken @kananyi]

The demo https://t.co/s5WZyJWtXb has now been run from Australia

[Ken @kananyi]

It’s my birthday. Instead of gifts, I want data. I just shared something I’ve been building since the start of this year. The full story is here. And I need your help proving there’s a Nigerian in every country on earth.