Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

Jev is actually insane. We benchmarked it, and the results completely change the game for us: > 500 real-time agents > running in parallel > in a 3D environment The preliminary results are crazy: > 500ms average latency > 35 API calls/s > all with a naive implementation We... show more

Cris Lenta

2,154 subscribers

58,675 Aufrufe • vor 3 Tagen •via X (Twitter)

Bildung Gesundheit & Wellness Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

21 Kommentare

Profilbild von PSYMOD

PSYMODvor 3 Tagen

Working on real world OSM reasoning and decision making with JEV ! CRAZY OUTCOMES

Profilbild von Cris Lenta

Cris Lentavor 3 Tagen

only a preview whats coming will be insane

Profilbild von PG

PGvor 3 Tagen

Maybe this is the start of an opening into micro-bot biological monitoring / repairs though

Profilbild von Cris Lenta

Cris Lentavor 3 Tagen

tell me more

Profilbild von Dae

Daevor 3 Tagen

Cooooooool

Profilbild von Cris Lenta

Cris Lentavor 3 Tagen

v coooooool

Profilbild von Alexander Goslin

Alexander Goslinvor 3 Tagen

This was one of the first applications that came to mind for Jev. Super cool!

Profilbild von Cris Lenta

Cris Lentavor 3 Tagen

maybe we should work together

Profilbild von PG

PGvor 3 Tagen

Don't start making real-life nanobot swarms 😑

Profilbild von Cris Lenta

Cris Lentavor 3 Tagen

no nano only real-size

Profilbild von Sebastiano Mandalà

Sebastiano Mandalàvor 3 Tagen

They shouldn't release insane AI, so dangerous

Profilbild von Michael Waitze

Michael Waitzevor 3 Tagen

The latency improvements for parallel agents are going to be a total game changer. We actually went deeper on this here:

Profilbild von CuriousDev

CuriousDevvor 3 Tagen

Isn't this close to the state space models, we can remember manipulate and update states this definitely looks interesting

Profilbild von Sacha Martinelle

Sacha Martinellevor 3 Tagen

Impressive! How does that compare to the previous top model?

Profilbild von A Eye Bubble

A Eye Bubblevor 3 Tagen

No you're insane for falling for this scam. Hahahshaha! You're dum

Profilbild von Geωrge

Geωrgevor 3 Tagen

so basically now we have the haves and havenots in low latency token generation

Profilbild von Gregor

Gregorvor 3 Tagen

500ms average is the easy number for multi-agent. genuinely asking what does p99 look like when all 500 agents cluster in the same region at once?

Profilbild von mblaso

mblasovor 3 Tagen

@crislenta what’s insane is that you haven’t posted it on

Profilbild von dp

dpvor 3 Tagen

Probably great for game agents

Profilbild von Justin He

Justin Hevor 3 Tagen

Because JEV is type-safe, doesn't this have limitations on the choices and directions these simulation steer towards? aka not completely free-will for each agent (lol)

Profilbild von AI News Daily

AI News Dailyvor 3 Tagen

Parallel agents are impressive, but the quality checks will decide how useful this is. I would love to see the error rate under real workloads.

Ähnliche Videos

The headshot animations seem to be faster. Tests need to be replicated more for confirmation. This is with loopback=0. With latency on a server results may vary more. (Not a fact :/ it's a theory)

CS2 Kitchen

109,804 Aufrufe • vor 8 Monaten

The real bottleneck for AI agents? It's not the model. It's search. Even a top-tier LLM will struggle if it can't access reliable, up-to-date info. I recently tested AnySearch in my agent workflow. Right away, the results were SO much easier to handle. Let's dive in 🧵↓

The real bottleneck for AI agents? It's not the model. It's search. Even a top-tier LLM will struggle if it can't access reliable, up-to-date info. I recently tested AnySearch in my agent workflow. Right away, the results were SO much easier to handle. Let's dive in 🧵↓

Charly Wargnier ♨️

16,744 Aufrufe • vor 2 Monaten

Introducing /search - the simplest way for agents and devs to discover the web 👀 Our #1 most requested endpoint is finally here. Search the web AND scrape all results in an LLM-ready format with one API call. Now live on our API, MCP, and all integrations 🔥

Introducing /search - the simplest way for agents and devs to discover the web 👀 Our #1 most requested endpoint is finally here. Search the web AND scrape all results in an LLM-ready format with one API call. Now live on our API, MCP, and all integrations 🔥

Firecrawl

390,772 Aufrufe • vor 1 Jahr

Wow, TypeSafe AI Jev is a game changer for voice-driven UI. Responsiveness really matters for an experience to 'feel' right, and this certainly does the job! Latency here is not optimal given I'm in London and calling out to us-west. But still, super snappy. In this video, Jamcat is using Pipecat to orchestrate a real-time handover between a command agent (Jev) that drives the UI, and voice agent (PhoneLLM) for handling generations and (not in the video) discussing ideas and session setup with the user.

Wow, TypeSafe AI Jev is a game changer for voice-driven UI. Responsiveness really matters for an experience to 'feel' right, and this certainly does the job! Latency here is not optimal given I'm in London and calling out to us-west. But still, super snappy. In this video, Jamcat is using Pipecat to orchestrate a real-time handover between a command agent (Jev) that drives the UI, and voice agent (PhoneLLM) for handling generations and (not in the video) discussing ideas and session setup with the user.

Jon Taylor

43,880 Aufrufe • vor 3 Tagen

Joe Rogan says Trump is not the answer in 2024: “Here's the real problem. Republicans won't be the solution either, kids. The problem is people in a position of power. The Republicans seem like they are your solution, but it's just because the people in power right now are the Democrats. Whenever the Democrats are out and the Republicans are in, everybody is dying for a Democrat.” “It's just the same structure with a different face.” It’s crazy how short people’s memories are. The Biden administration has been such a clown show that people now believe Trump and a red wave will be a solution to their problems and disregard what it was like while Trump was in office. Then in 2028 after another Trump clown show people will be hoping for a rational middle of the road Democrat and blue wave. The cycle is never ending. The only way to break the cycle is to change the system. The only way to change the system is we start voting with our hearts not party. We start voting with a candidate we most align with and completely disregard what political party they designate to the side of their name. The only way we change anything is if we all individually take the leap in 2024 and vote for someone we truly believe in.

Joe Rogan says Trump is not the answer in 2024: “Here's the real problem. Republicans won't be the solution either, kids. The problem is people in a position of power. The Republicans seem like they are your solution, but it's just because the people in power right now are the Democrats. Whenever the Democrats are out and the Republicans are in, everybody is dying for a Democrat.” “It's just the same structure with a different face.” It’s crazy how short people’s memories are. The Biden administration has been such a clown show that people now believe Trump and a red wave will be a solution to their problems and disregard what it was like while Trump was in office. Then in 2028 after another Trump clown show people will be hoping for a rational middle of the road Democrat and blue wave. The cycle is never ending. The only way to break the cycle is to change the system. The only way to change the system is we start voting with our hearts not party. We start voting with a candidate we most align with and completely disregard what political party they designate to the side of their name. The only way we change anything is if we all individually take the leap in 2024 and vote for someone we truly believe in.

End Tribalism in Politics

2,566,509 Aufrufe • vor 2 Jahren

$10 Cardano is realistic. And I'm not crazy... I did not expect this at all! I was testing our AI in the CCV Intelligence system, & asking for $ADA exit strategy. It gave me a $10 Cardano peak, which is great. But it's the reasoning that completely shocked me! Watch this now!

$10 Cardano is realistic. And I'm not crazy... I did not expect this at all! I was testing our AI in the CCV Intelligence system, & asking for $ADA exit strategy. It gave me a $10 Cardano peak, which is great. But it's the reasoning that completely shocked me! Watch this now!

Dan Gambardello

73,051 Aufrufe • vor 1 Jahr

All of this crazy in one week. They are overwhelming the system and keeping us in a loop. This is not a guess.

All of this crazy in one week. They are overwhelming the system and keeping us in a loop. This is not a guess.

MAGA Cult Slayer

14,749 Aufrufe • vor 8 Monaten

Okey, now this is wild. The latency is crazy fast, and I’m not sure we are ready for this kind of machine interactions. Best part you can try it yourself now link below 👇🏼

Okey, now this is wild. The latency is crazy fast, and I’m not sure we are ready for this kind of machine interactions. Best part you can try it yourself now link below 👇🏼

Linus ✦ Ekenstam

224,497 Aufrufe • vor 2 Jahren

BREAKING: Did Polymarket just silently nuke the 500ms taker delay? Rumor has it the speed bump on crypto markets is GONE. No announcement, no changelog, nothing. The crazy part? It was never officially documented in the first place. I literally only found proof of its existence buried in a 3rd-party GitHub repo code Why this matters: - Takers/aggressive bots: You're eating MUCH faster now, which is great... if your latency is good - Makers/liquidity providers: You're about to get picked off way more often. RIP to anyone not updating quotes constantly - Arbitrage: The game just got 500ms faster. May the lowest-latency chad win Traders are already claiming they got wrecked overnight. Unless you asked Claude to update your OpenClaw configs for zero-latency, you are probably getting farmed right now See the execution speed for yourself:

BREAKING: Did Polymarket just silently nuke the 500ms taker delay? Rumor has it the speed bump on crypto markets is GONE. No announcement, no changelog, nothing. The crazy part? It was never officially documented in the first place. I literally only found proof of its existence buried in a 3rd-party GitHub repo code Why this matters: - Takers/aggressive bots: You're eating MUCH faster now, which is great... if your latency is good - Makers/liquidity providers: You're about to get picked off way more often. RIP to anyone not updating quotes constantly - Arbitrage: The game just got 500ms faster. May the lowest-latency chad win Traders are already claiming they got wrecked overnight. Unless you asked Claude to update your OpenClaw configs for zero-latency, you are probably getting farmed right now See the execution speed for yourself:

BuBBliK

110,516 Aufrufe • vor 7 Monaten

Over half of Californians think our state is on the wrong track, and the number is only growing. We pay the highest taxes in the country just to get the worst results. Not for long! We will deliver real change in November!

Over half of Californians think our state is on the wrong track, and the number is only growing. We pay the highest taxes in the country just to get the worst results. Not for long! We will deliver real change in November!

Steve Hilton

31,939 Aufrufe • vor 5 Monaten

For the first time in The Associated Press 's history, we will offer a digital livestream of Tuesday night's US presidential election results. I'm honored to be the anchor of this inaugural broadcast that will include expert discussions, live vote counts, rallies, and results from over 35 cameras nationwide. Get real-time AP VoteCast survey results and live electoral map updates every 30 minutes.

For the first time in The Associated Press 's history, we will offer a digital livestream of Tuesday night's US presidential election results. I'm honored to be the anchor of this inaugural broadcast that will include expert discussions, live vote counts, rallies, and results from over 35 cameras nationwide. Get real-time AP VoteCast survey results and live electoral map updates every 30 minutes.

Meg Kinnard

101,330 Aufrufe • vor 1 Jahr

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

kwindla

330,647 Aufrufe • vor 24 Tagen

🚨 This is Warzone on the highest fps/ lowest latency settings 🚨 It is really hard to lose a straight up gunfight when you are on this low of latency. Feels similar to the "host advantage" we all used to want back in the console days.

🚨 This is Warzone on the highest fps/ lowest latency settings 🚨 It is really hard to lose a straight up gunfight when you are on this low of latency. Feels similar to the "host advantage" we all used to want back in the console days.

Kibbs

301,424 Aufrufe • vor 2 Jahren

$TESTED: Windows 11's upcoming "Low Latency Profile" mode brings genuine performance improvements to the OS, speeding up flyout and app launches significantly. We've benchmarked opening some apps on video with the Low Latency Profile enabled and disabled, and you can see differences in how quickly things appear. For some things, it's a fraction of a second faster, for others, it's a significant increase in speed. In our testing, this new Low Latency Profile is a major improvement in overall responsiveness when it comes to opening apps and flyouts. Our tests were conducted on a clean install of the latest Windows 11 preview build on the same hardware.$

TESTED: Windows 11's upcoming "Low Latency Profile" mode brings genuine performance improvements to the OS, speeding up flyout and app launches significantly. We've benchmarked opening some apps on video with the Low Latency Profile enabled and disabled, and you can see differences in how quickly things appear. For some things, it's a fraction of a second faster, for others, it's a significant increase in speed. In our testing, this new Low Latency Profile is a major improvement in overall responsiveness when it comes to opening apps and flyouts. Our tests were conducted on a clean install of the latest Windows 11 preview build on the same hardware.

Windows Central

1,162,358 Aufrufe • vor 4 Monaten

another example: a developer used the exact same $8 ESP32-S3 chip and local AI setup to build a real-time object recognition camera the fact that this is running entirely offline, on an eight-dollar microcontroller, without any cloud API calls, is wild... it captures the video feed, processes it locally through the model, and identifies objects in real-time. we are entering an era where you can just glue intelligence onto any physical object for less than $10. the hardware barrier is completely shattered.

another example: a developer used the exact same $8 ESP32-S3 chip and local AI setup to build a real-time object recognition camera the fact that this is running entirely offline, on an eight-dollar microcontroller, without any cloud API calls, is wild... it captures the video feed, processes it locally through the model, and identifies objects in real-time. we are entering an era where you can just glue intelligence onto any physical object for less than $10. the hardware barrier is completely shattered.

ard

225,265 Aufrufe • vor 1 Monat

Mikal Bridges is asked about the Knicks' playoff winning streak: "We don't care too much about the winning streak. All that's in the past. To us, it's 0-0. Take it one game at a time. Our most important game is the next game."

Mikal Bridges is asked about the Knicks' playoff winning streak: "We don't care too much about the winning streak. All that's in the past. To us, it's 0-0. Take it one game at a time. Our most important game is the next game."

SNY Knicks

40,174 Aufrufe • vor 3 Monaten

The engine of the 21st century is here. 🧠 The QVAC SDK is the "steam engine" of the AI era—decoupling intelligence from the cloud and putting it in your hands. A single API for local-first, modular AI that runs anywhere. - Sovereign: Own your engine, don't rent it. - Local: 0 latency, no cloud dependency. - Modular: Stackable, universal building blocks. The era of Stable Intelligence has begun.

The engine of the 21st century is here. 🧠 The QVAC SDK is the "steam engine" of the AI era—decoupling intelligence from the cloud and putting it in your hands. A single API for local-first, modular AI that runs anywhere. - Sovereign: Own your engine, don't rent it. - Local: 0 latency, no cloud dependency. - Modular: Stackable, universal building blocks. The era of Stable Intelligence has begun.

QVAC

10,663,755 Aufrufe • vor 5 Monaten

The new damage prediction system in CS2 is absolutely amazing under 30 ish ping. The game feels insane to play and the kill delay i used to feel has completely vanished. The headshots are 10 times more satisfying with this implementation.

The new damage prediction system in CS2 is absolutely amazing under 30 ish ping. The game feels insane to play and the kill delay i used to feel has completely vanished. The headshots are 10 times more satisfying with this implementation.

elias

373,484 Aufrufe • vor 1 Jahr