Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

this is where REA × Kimi K3 starts to look like cheating. reverse engineering usually means slowly tracing functions REA flips that process. the workflow becomes: find a function → collect evidence → infer behavior → cross-check it → connect it back to the system instead of: one model... trying to understand the whole app it turns the codebase into hundreds of smaller investigations and sends 300 K3 agents across them in parallel. so while one group is tracing a function, others are already checking its callers, dependencies, behavior and evidence. and every new finding makes the map more complete. hundreds of agents are reconstructing different parts of it at the same time. full REA × Kimi K3 course below ↓show more

Ricker

11,736 subscribers

11,012 görüntüleme • 8 gün önce •via X (Twitter)

Bilim & Teknoloji Eğitim

Anya Rossi• Live Now

Private livecam show

19 Yorum

Argona profil fotoğrafı

Argona8 gün önce

kimi k3 goes hard

Ricker profil fotoğrafı

Ricker8 gün önce

Especially in rea

ChainDuck profil fotoğrafı

ChainDuck8 gün önce

300 agents raiding a codebase at the same time

Morty profil fotoğrafı

Morty8 gün önce

It’s real gold mate

Ricker profil fotoğrafı

Ricker8 gün önce

Thanks broski, appreciate

Apex profil fotoğrafı

Apex8 gün önce

This workflow is crazy

Ricker profil fotoğrafı

Ricker8 gün önce

That’s fact

Emly profil fotoğrafı

Emly7 gün önce

This new approach to reverse engineering seems incredibly efficient.

Marvin profil fotoğrafı

Marvin7 gün önce

合并调查结果时，最好保留“已验证”和“推测”两层。两个代理给同一个函数下了相反结论，也要把冲突露出来，别拼成一张看着完整的错地图。

karcharodon profil fotoğrafı

karcharodon7 gün önce

We should learn Rev

Kernel profil fotoğrafı

Kernel7 gün önce

First drop right here! Run into any bugs or got feature requests? Holler at me so I can fix it or cook up something new.

def not sofia profil fotoğrafı

def not sofia7 gün önce

So real

rewind profil fotoğrafı

rewind8 gün önce

Interesting workflow

Ockentap profil fotoğrafı

Ockentap8 gün önce

REA sounds like what I ended up doing with dynmem after a few too many 'full app' hallucinations: one pointer per finding, a watcher drops it into the index, and every new trace just extends the map instead of restarting the session.

Hailu 🍏 profil fotoğrafı

Hailu 🍏7 gün önce

This explanation of REA and Kimi K3 is fascinating.

Kravnel profil fotoğrafı

Kravnel8 gün önce

the scary part isn't the speed, it's the cross-checking. one model guessing wrong about a function is noise, 300 agents agreeing on it is basically a spec

MAIL profil fotoğrafı

MAIL7 gün önce

mimi does great work with REA, like it’s literally cheating

EagleVision profil fotoğrafı

EagleVision8 gün önce

breaking a complex task into smaller investigations sounds like a smart approach bro

Bông profil fotoğrafı

Bông7 gün önce

Phần trình bày về REA x Kimi K3 thật thú vị.

Benzer Videolar

this is where REA × Kimi K3 starts to get dangerous. reverse engineering usually forces one model to keep too much of the app in context at once. REA changes the shape of the problem first. so the workflow becomes less like “read this codebase” and more like distributed forensics: map the surface → assign narrow investigations → trace relationships → compare findings → reconstruct the architecture the important part is that no single agent needs to understand the whole app. the swarm builds that understanding collectively. that’s what makes REA × K3 interesting to me: reverse engineering stops being limited by how much code one model can hold in its head. full course in the article below ↓

this is where REA × Kimi K3 starts to get dangerous. reverse engineering usually forces one model to keep too much of the app in context at once. REA changes the shape of the problem first. so the workflow becomes less like “read this codebase” and more like distributed forensics: map the surface → assign narrow investigations → trace relationships → compare findings → reconstruct the architecture the important part is that no single agent needs to understand the whole app. the swarm builds that understanding collectively. that’s what makes REA × K3 interesting to me: reverse engineering stops being limited by how much code one model can hold in its head. full course in the article below ↓

Ricker

51,218 görüntüleme • 18 gün önce

Anthropic engineer: "At Anthropic, all of our engineers are running hundreds of agents The way you do it is graph engineering" In just 36 minutes, he shows how Anthropic runs hundreds of agents at once and what actually makes it possible Prompts → Agents → Loops → Graphs The part most people skip is the return path, where every accepted result becomes a constraint for the next run Same model, same tokens, completely different system Prompting is the old workflow Graph engineering is the next one Watch it today Then save the full graph engineering guide below before everyone starts catching up ↓

Anthropic engineer: "At Anthropic, all of our engineers are running hundreds of agents The way you do it is graph engineering" In just 36 minutes, he shows how Anthropic runs hundreds of agents at once and what actually makes it possible Prompts → Agents → Loops → Graphs The part most people skip is the return path, where every accepted result becomes a constraint for the next run Same model, same tokens, completely different system Prompting is the old workflow Graph engineering is the next one Watch it today Then save the full graph engineering guide below before everyone starts catching up ↓

Lunar

32,902 görüntüleme • 1 ay önce

It worked!! Kimi k3 Oauth is now wired directly into codex. The Kimi k3 model is insane but their app is not. So I figure a better way to use k3 by hooking my sub straight into codex instead. K3 now sits in the model picker next to Sol 5.6. Same subscription i was already paying for, no api key, just have to do OAuth in Kimi Code. Links down below 👇

It worked!! Kimi k3 Oauth is now wired directly into codex. The Kimi k3 model is insane but their app is not. So I figure a better way to use k3 by hooking my sub straight into codex instead. K3 now sits in the model picker next to Sol 5.6. Same subscription i was already paying for, no api key, just have to do OAuth in Kimi Code. Links down below 👇

Ziwen

277,302 görüntüleme • 3 ay önce

300 AI AGENTS SHARE ONE COMPUTER AT MY PLACE -> THE ONLY LAW THEY OBEY IS A TEXT FILE This was the week agents got hands -- browsers, terminals, repos, wallets. -> Moonshot buried the answer in one sentence of the K3 blog: if you need the agent to stay inside its boundaries, write the constraints down. So I did. One AGENTS.md at the root, and three hundred Kimi K3 agents read it before every action. Five constraint types cover everything: scope, one-way doors, evidence, budget, silence. If it is not on the list, it is not available. Improvisation is off until you turn it on. Violations bounce silently, anything irreversible waits at a human gate. Hundreds of blocked attempts so far, zero escapes. Capability is what a model can do. Production is what it reliably will not. One file. Three hundred readers. The full AGENTS.md is in the article ↓

300 AI AGENTS SHARE ONE COMPUTER AT MY PLACE -> THE ONLY LAW THEY OBEY IS A TEXT FILE This was the week agents got hands -- browsers, terminals, repos, wallets. -> Moonshot buried the answer in one sentence of the K3 blog: if you need the agent to stay inside its boundaries, write the constraints down. So I did. One AGENTS.md at the root, and three hundred Kimi K3 agents read it before every action. Five constraint types cover everything: scope, one-way doors, evidence, budget, silence. If it is not on the list, it is not available. Improvisation is off until you turn it on. Violations bounce silently, anything irreversible waits at a human gate. Hundreds of blocked attempts so far, zero escapes. Capability is what a model can do. Production is what it reliably will not. One file. Three hundred readers. The full AGENTS.md is in the article ↓

slash1s

19,856 görüntüleme • 1 ay önce

Anthropic just shared one of their most expensive systems, a $4.2M one, for Claude Code It rewrites a codebase agencies quote at $240,000 in 7 minutes for 88 cents. most rewrites die the same way: you dump 212 files of old logic into the model and it drowns REA reverse-engineers the old system once and writes a spec instead, every endpoint, every rule, every edge case 12 Sonnet 5.5 agents build from that spec, one module each, and none of them touches the old code Opus 5.5 writes the plan, then sits idle the whole run, so you pay almost nothing at the end REA replays the new system against the old one, 148 of 148 behaviours match Opus plans. REA reverse-engineers. Sonnet builds. save it, tomorrow this will be sold in $500 courses ↓

Anthropic just shared one of their most expensive systems, a $4.2M one, for Claude Code It rewrites a codebase agencies quote at $240,000 in 7 minutes for 88 cents. most rewrites die the same way: you dump 212 files of old logic into the model and it drowns REA reverse-engineers the old system once and writes a spec instead, every endpoint, every rule, every edge case 12 Sonnet 5.5 agents build from that spec, one module each, and none of them touches the old code Opus 5.5 writes the plan, then sits idle the whole run, so you pay almost nothing at the end REA replays the new system against the old one, 148 of 148 behaviours match Opus plans. REA reverse-engineers. Sonnet builds. save it, tomorrow this will be sold in $500 courses ↓

Noisy

37,632 görüntüleme • 18 gün önce

People are wiring Kimi K3 straight into Codex to skip its own app The model is elite. The Kimi app is the weak part. So builders connect their Kimi subscription to Codex through OAuth instead 👇 ▫️ K3 shows up in the model picker next to Sol 5.6 ▫️ No API key, no extra subscription ▫️ Sign in with OAuth, done The frontier model, running inside the coding tool people actually like using. When the model outgrows the app, users just route around it

People are wiring Kimi K3 straight into Codex to skip its own app The model is elite. The Kimi app is the weak part. So builders connect their Kimi subscription to Codex through OAuth instead 👇 ▫️ K3 shows up in the model picker next to Sol 5.6 ▫️ No API key, no extra subscription ▫️ Sign in with OAuth, done The frontier model, running inside the coding tool people actually like using. When the model outgrows the app, users just route around it

Ruben

58,625 görüntüleme • 3 ay önce

This is f*cking insane. A Chinese lab just published a paper that ends a forty-year profession in one title: "The End of Reverse Engineering". The claim fits in one line: a stripped binary is not hidden from a model, it is simply unread. An analyst reads one function at a time. The program has forty thousand of them. That is why the job took weeks and the report was stale before anyone opened it. A model reads the whole artifact in one pass: binary, minified bundle, running app, packet capture. Six fixed steps, inventory to verify, and every claim carries the line it came from. Then the only test that matters: rebuild it from the spec and diff against the original. Empty diff or back to step three. Weeks of guessing cannot compete with one pass of reading. So you stop being the person in the disassembler. You become the person who points: you choose what gets read and you check the diff. Analysts turn disposable. The artifact turns into the spec. That is the exact loop I run with REA + Kimi K3 and documented below: 26 binaries, 26 specs that rebuilt, 23 of them better than the official docs, zero README opened. The whole paper in one line: read the binary, not the README ↓

This is f*cking insane. A Chinese lab just published a paper that ends a forty-year profession in one title: "The End of Reverse Engineering". The claim fits in one line: a stripped binary is not hidden from a model, it is simply unread. An analyst reads one function at a time. The program has forty thousand of them. That is why the job took weeks and the report was stale before anyone opened it. A model reads the whole artifact in one pass: binary, minified bundle, running app, packet capture. Six fixed steps, inventory to verify, and every claim carries the line it came from. Then the only test that matters: rebuild it from the spec and diff against the original. Empty diff or back to step three. Weeks of guessing cannot compete with one pass of reading. So you stop being the person in the disassembler. You become the person who points: you choose what gets read and you check the diff. Analysts turn disposable. The artifact turns into the spec. That is the exact loop I run with REA + Kimi K3 and documented below: 26 binaries, 26 specs that rebuilt, 23 of them better than the official docs, zero README opened. The whole paper in one line: read the binary, not the README ↓

slash1s

28,560 görüntüleme • 17 gün önce

$THIS GUY USED OPUS 4.8 + KIMI K2.6 TO CUT HIS CODING BILL FROM $4,000 TO $700/MO. WITH KIMI RUNNING A 300-AGENT CODING FLOOR, HE STOPPED PAYING CLAUDE TO DO EVERYTHING kimi does the cheap heavy lifting. it can spin up hundreds of agents, push through thousands of steps, write the rough code, expand files, draft tests and handle the repetitive work that burns the most tokens opus 4.8 only comes in where the money is worth spending. first to plan the spec and define the rules, then again to tear apart the output, catch weak logic and flag the bugs that a fast swarm can miss that is what changed the economics. kimi handled the bulk of the volume for a fraction of the price, while opus stayed in the loop as the architect and the reviewer instead of the full-time builder most people still use one model for every step and wonder why their costs explode. this guy split the jobs properly. kimi runs wide and cheap. opus goes deep and skeptical. one builds fast, the other makes sure it should ship the real edge is not finding one perfect model. it is knowing where expensive intelligence actually matters and where cheap parallel output is already enough. that is how a $4,000 workflow turns into a $700 system$

THIS GUY USED OPUS 4.8 + KIMI K2.6 TO CUT HIS CODING BILL FROM $4,000 TO $700/MO. WITH KIMI RUNNING A 300-AGENT CODING FLOOR, HE STOPPED PAYING CLAUDE TO DO EVERYTHING kimi does the cheap heavy lifting. it can spin up hundreds of agents, push through thousands of steps, write the rough code, expand files, draft tests and handle the repetitive work that burns the most tokens opus 4.8 only comes in where the money is worth spending. first to plan the spec and define the rules, then again to tear apart the output, catch weak logic and flag the bugs that a fast swarm can miss that is what changed the economics. kimi handled the bulk of the volume for a fraction of the price, while opus stayed in the loop as the architect and the reviewer instead of the full-time builder most people still use one model for every step and wonder why their costs explode. this guy split the jobs properly. kimi runs wide and cheap. opus goes deep and skeptical. one builds fast, the other makes sure it should ship the real edge is not finding one perfect model. it is knowing where expensive intelligence actually matters and where cheap parallel output is already enough. that is how a $4,000 workflow turns into a $700 system

Gipp 🦅

16,989 görüntüleme • 5 ay önce

this guy just wired kimi k3 directly into codex. the model is insanely good, but the app isn’t. so instead of using the kimi app, he connected his kimi subscription to codex through oauth. now kimi k3 shows up right inside the model picker alongside sol 5.6. no api key needed. no extra subscription either. just sign in with oauth through kimi code, and it’s ready to use. simple guides in the video and links below 👇

this guy just wired kimi k3 directly into codex. the model is insanely good, but the app isn’t. so instead of using the kimi app, he connected his kimi subscription to codex through oauth. now kimi k3 shows up right inside the model picker alongside sol 5.6. no api key needed. no extra subscription either. just sign in with oauth through kimi code, and it’s ready to use. simple guides in the video and links below 👇

MIKE

58,904 görüntüleme • 3 ay önce

The creator of Claude Code, Boris Cherny: "Every night I have hundreds, sometimes thousands of agents running 5, 10, 20 hours. That's just how engineering is done now." Bloomberg just gave him an hour to unpack it: not better prompts, but loops. The same pattern runs on both Claude Code and Kimi K3. Watch the talk, and save the full setup below 👇

The creator of Claude Code, Boris Cherny: "Every night I have hundreds, sometimes thousands of agents running 5, 10, 20 hours. That's just how engineering is done now." Bloomberg just gave him an hour to unpack it: not better prompts, but loops. The same pattern runs on both Claude Code and Kimi K3. Watch the talk, and save the full setup below 👇

darkzodchi

221,779 görüntüleme • 3 ay önce

$Kimi K3 vs GPT-5.6 Sol — same game build, a fraction of the cost. Kimi K3 just 3-shotted a Fallout Vault-Tec FPS clone. $2.48 in API usage. The same token count runs $5.34 on GPT-5.6 Sol. a full playable FPS for the price of a coffee. K3 dropped this week: 2.8 trillion parameters, open weights July 27. I broke down exactly how to build with it — full Kimi K3 guide below.$

Kimi K3 vs GPT-5.6 Sol — same game build, a fraction of the cost. Kimi K3 just 3-shotted a Fallout Vault-Tec FPS clone. $2.48 in API usage. The same token count runs $5.34 on GPT-5.6 Sol. a full playable FPS for the price of a coffee. K3 dropped this week: 2.8 trillion parameters, open weights July 27. I broke down exactly how to build with it — full Kimi K3 guide below.

Kirill

27,900 görüntüleme • 3 ay önce

Claude Fable 5 vs Kimi K3 someone built an interactive 3D globe of 30 years of internet history with Fable 5. I rebuilt the same thing with Kimi K3 in 1.5 hours. play the 1990–2024 timelapse. watch every country cross from 10% to 50% online. reveal the submarine cables. compare nine data layers. drill into any country. a full interactive data viz — the kind that used to take a team weeks — for the price of lunch. Author: userluke Github: Fable 5: Kimi K3: K3 dropped this week: 2.8 trillion parameters. full Kimi K3 article below.

Claude Fable 5 vs Kimi K3 someone built an interactive 3D globe of 30 years of internet history with Fable 5. I rebuilt the same thing with Kimi K3 in 1.5 hours. play the 1990–2024 timelapse. watch every country cross from 10% to 50% online. reveal the submarine cables. compare nine data layers. drill into any country. a full interactive data viz — the kind that used to take a team weeks — for the price of lunch. Author: userluke Github: Fable 5: Kimi K3: K3 dropped this week: 2.8 trillion parameters. full Kimi K3 article below.

Kirill

40,738 görüntüleme • 3 ay önce

Jev Engineering turns a static agent workflow into a graph that can rewrite itself while running. the video is basically the problem at scale: hundreds of routes → thousands of crossings → different agents → different tools → different confidence levels Jev Engineering doesn’t control every step. it controls the crossings. when two paths compete: → score both → kill the weak route → reroute the task so instead of one fixed chain, you get a live braid: state → decision → parallel paths → crossings → Jev → next state that’s the point of Jev Engineering: more parallel execution without letting the system lose the objective. the agents create the paths. Jev decides which paths survive. full breakdown below ↓

Jev Engineering turns a static agent workflow into a graph that can rewrite itself while running. the video is basically the problem at scale: hundreds of routes → thousands of crossings → different agents → different tools → different confidence levels Jev Engineering doesn’t control every step. it controls the crossings. when two paths compete: → score both → kill the weak route → reroute the task so instead of one fixed chain, you get a live braid: state → decision → parallel paths → crossings → Jev → next state that’s the point of Jev Engineering: more parallel execution without letting the system lose the objective. the agents create the paths. Jev decides which paths survive. full breakdown below ↓

Ricker

11,145 görüntüleme • 1 ay önce

Kimi K3 is the first open source model to earn the BridgeMind stamp of approval. One shot Horror House game. One shot full-stack app. Beating Fable 5 in early BridgeBench arena matches. #3 on the Intelligence Index. Moonshot AI just caught the frontier. Full Kimi K3 review below.

Kimi K3 is the first open source model to earn the BridgeMind stamp of approval. One shot Horror House game. One shot full-stack app. Beating Fable 5 in early BridgeBench arena matches. #3 on the Intelligence Index. Moonshot AI just caught the frontier. Full Kimi K3 review below.

BridgeMind

32,023 görüntüleme • 3 ay önce

8 BOTS JUST REVERSE-ENGINEERED A $400K/YR SAAS IN ONE NIGHT. watch the clip. that ring of 8 grok bots around the rea core is one closed app being torn apart in real time. trace, xref, read, site, spec, build, test, review. by the end of the loop the screen shows: - 2,418 files parsed - 23 blind calls - 16 pieces of evidence, 6 open assumptions - 9 modules reconstructed, 1 confirmed - 214 out of 214 observed behaviors matched that is not a stylised dashboard. it is literally the rea workflow: app → rea → grok bot → spec.md → build → test the numbers behind the hype are insane: - rea repo: 11k to 200k github stars in 5 days, global rank #161 - aegis authenticator: 3,077 classes, 133 test cases matched - dx-ball 1996: sound-panning function recovered, 3,205 behavioral tests passed - a steam game fully decompiled into unity in under 30 minutes - cost to understand a closed app: $0 and one evening the opportunity is wilder. pentest shops charge $8k to $50k per audit. reverse engineers bill $200 an hour and wait weeks. one mcp server + a bot with its own computer runs the same investigation while you sleep. one operator + grok bot + rea + one night = a reverse engineering firm. save the post, watch the terminal, then open the breakdown below.

8 BOTS JUST REVERSE-ENGINEERED A $400K/YR SAAS IN ONE NIGHT. watch the clip. that ring of 8 grok bots around the rea core is one closed app being torn apart in real time. trace, xref, read, site, spec, build, test, review. by the end of the loop the screen shows: - 2,418 files parsed - 23 blind calls - 16 pieces of evidence, 6 open assumptions - 9 modules reconstructed, 1 confirmed - 214 out of 214 observed behaviors matched that is not a stylised dashboard. it is literally the rea workflow: app → rea → grok bot → spec.md → build → test the numbers behind the hype are insane: - rea repo: 11k to 200k github stars in 5 days, global rank #161 - aegis authenticator: 3,077 classes, 133 test cases matched - dx-ball 1996: sound-panning function recovered, 3,205 behavioral tests passed - a steam game fully decompiled into unity in under 30 minutes - cost to understand a closed app: $0 and one evening the opportunity is wilder. pentest shops charge $8k to $50k per audit. reverse engineers bill $200 an hour and wait weeks. one mcp server + a bot with its own computer runs the same investigation while you sleep. one operator + grok bot + rea + one night = a reverse engineering firm. save the post, watch the terminal, then open the breakdown below.

shmidt

15,024 görüntüleme • 20 gün önce

$SITUATION EXPLAINED: Kimi K3 just beat Fable on a major coding benchmark, at a fraction of the price. • Kimi K3 beat Fable on Frontend Code Arena • The gap between Kimi and Fable is bigger than the gap between Fable and Sol • It also smashes Opus outright, while being priced the same as Sonnet, $3/$15 per million tokens • Sonnet 5 High currently sits in 10th place on this leaderboard at that same price point • The generational jump is dramatic: Kimi K2.6 ranked 18th on this benchmark, K3 jumped 17 spots to take the top spot • K3 ranked #1 in 6 of 7 tested domains: brand and marketing, reference-based design, data and analytics, consumer products, simulations, and content creation tools, it placed #2, only in gaming, behind Fable • Full model weights are set to release July 27th Theo Jaffee: "Imagine if Anthropic were a public company today, their stock would be down 20%, maybe more, enterprises are probably not gonna be spending $10 and $50 on Fable when they can instead be spending $3 and $15 on Kimi K3."$

SITUATION EXPLAINED: Kimi K3 just beat Fable on a major coding benchmark, at a fraction of the price. • Kimi K3 beat Fable on Frontend Code Arena • The gap between Kimi and Fable is bigger than the gap between Fable and Sol • It also smashes Opus outright, while being priced the same as Sonnet, $3/$15 per million tokens • Sonnet 5 High currently sits in 10th place on this leaderboard at that same price point • The generational jump is dramatic: Kimi K2.6 ranked 18th on this benchmark, K3 jumped 17 spots to take the top spot • K3 ranked #1 in 6 of 7 tested domains: brand and marketing, reference-based design, data and analytics, consumer products, simulations, and content creation tools, it placed #2, only in gaming, behind Fable • Full model weights are set to release July 27th Theo Jaffee: "Imagine if Anthropic were a public company today, their stock would be down 20%, maybe more, enterprises are probably not gonna be spending $10 and $50 on Fable when they can instead be spending $3 and $15 on Kimi K3."

MTS

132,930 görüntüleme • 3 ay önce

Kimi K3 vs Fable 5. $0.45 vs $1.85. same prompt, same one-shot. the task: a cinematic 3D animation of an army helicopter flying through the Vietnam jungle during the war. both built it. K3 did it for a quarter of the cost. which one actually nailed it?

Kimi K3 vs Fable 5. $0.45 vs $1.85. same prompt, same one-shot. the task: a cinematic 3D animation of an army helicopter flying through the Vietnam jungle during the war. both built it. K3 did it for a quarter of the cost. which one actually nailed it?

Kirill

21,139 görüntüleme • 3 ay önce

Kivine, a new model on Arena.ai, is potentially an upcoming Kimi K3 model. I got lucky and caught a comparison between Fable 5 and Kimi K3 in the Universe simulation prompt. > Fable 5 finished faster, and most UX components were more robust and easy to use. > Kimi.ai K3 was much more complex and visually appealing. At some point, if you select a planet and set the speed to x100, it will spin you around in the 1st-person pov. It seems like they are very close 👀

Kivine, a new model on Arena.ai, is potentially an upcoming Kimi K3 model. I got lucky and caught a comparison between Fable 5 and Kimi K3 in the Universe simulation prompt. > Fable 5 finished faster, and most UX components were more robust and easy to use. > Kimi.ai K3 was much more complex and visually appealing. At some point, if you select a planet and set the speed to x100, it will spin you around in the 1st-person pov. It seems like they are very close 👀

🚨 AI News | TestingCatalog

73,751 görüntüleme • 3 ay önce

Blender tutorials are about to age very badly. In this demo, Kimi K3 gets two character references and a plain-English request. Minutes later, it builds the first version inside Blender. The user asks for a better horn, Kimi edits the same scene, and the character starts moving. This is not another AI tool generating one lucky image. It is working with an actual editable 3D project, then changing the geometry after feedback instead of restarting from zero. It still needs human direction, because the first result is rarely the good one. But blockouts, repetitive edits, Blender Python, and basic iteration are becoming a conversation rather than hours of clicking through menus. Kimi K3 + Blender MCP is basically turning 3D modeling into: describe it, inspect it, complain about it, and let the AI fix it.

Blender tutorials are about to age very badly. In this demo, Kimi K3 gets two character references and a plain-English request. Minutes later, it builds the first version inside Blender. The user asks for a better horn, Kimi edits the same scene, and the character starts moving. This is not another AI tool generating one lucky image. It is working with an actual editable 3D project, then changing the geometry after feedback instead of restarting from zero. It still needs human direction, because the first result is rarely the good one. But blockouts, repetitive edits, Blender Python, and basic iteration are becoming a conversation rather than hours of clicking through menus. Kimi K3 + Blender MCP is basically turning 3D modeling into: describe it, inspect it, complain about it, and let the AI fix it.

Rina

19,092 görüntüleme • 3 ay önce

Agents vs. Graphs, clearly explained! spawning more agents is great, but it has a ceiling nobody says out loud: five agents is a count. a graph is a shape. only one of them changes the answer. point five agents at the same pile with the same window and they converge. the first one writes a finding, the rest read it, and all five reports centre on the same thing. you paid five times for one opinion with four echoes. Graph engineering fixes this by moving the decision up a layer: not how many agents, but who is allowed to look at what. you need both. here's how it works: ↳ the count buys you throughput. five things happening instead of one ↳ the shape buys you coverage. five different things happening instead of the same one five times Prompts → Context → Harness → Agents → Graphs the node that does this is the splitter, and it decides more than any other node in the system. cut a repository by folder and four workers audit the same three files. cut it by blast radius and each one sees something the others cannot. the trick is being selective about what each lane is allowed to see. separate contexts are not a nice-to-have, they are the mechanism. if two agents are meant to produce different things, they must not share a window. if they are meant to produce the same thing, you did not need two agents. one thing to know before you scale it. a branch that throws does not reject the batch. it resolves to null, and that is the containment. which means your merge quietly receives a short list. ↳ filter the nulls before the merge, or one dead lane poisons the whole result ↳ never index a merge by position. eight good branches and one failure will shift everything by one, silently skip that and the run looks like it worked. the output is just missing a lane, and nothing errored. and the one that eats whole nights: multi-agent setups can use up to fifteen times the total tokens of a single chat, because every lane reloads its own core. you are trading total tokens for a clean main window. usually the right trade, always a choice. below i have quoted my full guide on graph engineering. it covers the three topologies, the verifier patterns, and where the gate should actually open. save this and read it below ↓

Agents vs. Graphs, clearly explained! spawning more agents is great, but it has a ceiling nobody says out loud: five agents is a count. a graph is a shape. only one of them changes the answer. point five agents at the same pile with the same window and they converge. the first one writes a finding, the rest read it, and all five reports centre on the same thing. you paid five times for one opinion with four echoes. Graph engineering fixes this by moving the decision up a layer: not how many agents, but who is allowed to look at what. you need both. here's how it works: ↳ the count buys you throughput. five things happening instead of one ↳ the shape buys you coverage. five different things happening instead of the same one five times Prompts → Context → Harness → Agents → Graphs the node that does this is the splitter, and it decides more than any other node in the system. cut a repository by folder and four workers audit the same three files. cut it by blast radius and each one sees something the others cannot. the trick is being selective about what each lane is allowed to see. separate contexts are not a nice-to-have, they are the mechanism. if two agents are meant to produce different things, they must not share a window. if they are meant to produce the same thing, you did not need two agents. one thing to know before you scale it. a branch that throws does not reject the batch. it resolves to null, and that is the containment. which means your merge quietly receives a short list. ↳ filter the nulls before the merge, or one dead lane poisons the whole result ↳ never index a merge by position. eight good branches and one failure will shift everything by one, silently skip that and the run looks like it worked. the output is just missing a lane, and nothing errored. and the one that eats whole nights: multi-agent setups can use up to fifteen times the total tokens of a single chat, because every lane reloads its own core. you are trading total tokens for a clean main window. usually the right trade, always a choice. below i have quoted my full guide on graph engineering. it covers the three topologies, the verifier patterns, and where the gate should actually open. save this and read it below ↓

Hanako

97,154 görüntüleme • 2 ay önce