Generate brand videos within 5 minutes using AI
Joined May 2024
- Tweets 130
- Following 73
- Followers 36
- Likes 229
1 Photos and videos
Pinned Tweet
20 Aug 2025
Yes, our first feature! Many more to comeβ¦.
19 Aug 2025
π¨ MAJOR NEWS! π¨
@PavePilotAI just got featured in @SJMAJewelryNews β America's longest-running jewelry publication! Thank you SJN team for the support!
Our AI video creation platform is officially changing the game for jewelry marketing. Link in comment.
1
358
Jun 11
claude desktop spawns a 1.8GB VM just to chat (351 HN points). most AI tools are way too heavy for what they actually do
went the other way with PavePilot. one product photo in, finished video out in 60 seconds. zero setup. try it free
23
Jun 10
my actual AI stack (not another 50-tool roundup):
β claude (shipping writing)
β cursor (where the code lives)
β make (gluing apis together)
β fable 5 dropped yesterday β already testing
β opencv 5 (707 HN pts, vision pipelines)
6 tools. the rest is noise.
55
Jun 9
anthropic: $965B valuation, $47B run rate, filed IPO before stripe. if you had to put real money on it β does a $1T AI company hold up or is this 1999 with better pr?
50
Jun 8
agent era isn't hype anymore:
β karpathy treats AI like a coworker
β 24yo making $63K/mo from whatsapp agents
β agencies swapping staff for agent pods
β deepseek v4 beat gpt-5.5 on precision
β msft scout runs agents autonomously
strongest signal?
42
Jun 7
a teenager with no math degree cracked an erdΕs problem with chatgpt. the approach was so weird a stanford mathematician compared it to finding a chess opening no human would imagine. 80-year conjectures don't fall to brute force. they fall to new ways of seeing.
19
Jun 5
everyone's hyping the OpenAI IPO. the real story? Anthropic filed first at $965B. OpenAI isn't leading this race β they're chasing it. 65x revenue at $850B is a lot of faith
11
Jun 3
5 google I/O wins builders should care about:
β omni: any input β video
β 3.5 flash beats 3.1 pro at flash speed
β search agents that monitor the web 24/7
β antigravity: agent dev platform
β search builds mini apps on the fly
model wars might be over
69
Jun 2
3 AI giga-IPOs this year. $4T in market cap. openai loses $14B. anthropic burns $5.2B. who in AI is actually turning a profit in 2026? genuinely curious because i keep looking and can't find many
20
Jun 1
the internet is getting rebuilt for agents
β AWS decoupled compute/storage for agent bursts
β 31% of HTTP traffic = bots (Cloudflare)
β machine > human traffic by 2027
β agents SSHing into their own VMs now
built for clicks β built for agents
34
May 31
best AI agent user isn't a senior engineer. it's a dispatcher who can't code. brethorst proved it (333 HN pts): agents flip who ships. domain experts build now. engineers in unfamiliar domains can't tell plausible wrong from right. your domain knowledge is probably your only moat
17
May 30
"mcp is dead" and "AI is frontend's lost decade all over again" both trending on HN. same story β tooling that eats 10% of your context and deskills you isn't progress, it's 2015 react with a new coat of paint
11
May 29
"various llm smells" hit 247 HN pts. guy AI-polished his math blog then spotted the same phrasing everywhere. tell: "X is the Y of Z" constructions. you write them without noticing. Opus 4.8 just dropped. more AI writing coming. edge goes to people who delete their own AI output
23
May 28
simon willison says openai anthropic hit PMF in april. his bill: $2,180 in tokens for $200. enterprises? they pay full API price now and their LLM costs are going vertical. coding agents are so useful that companies just... pay. that's the whole signal
24
May 27
"use boring languages with LLMs" just hit 187 HN pts. LLMs amplify whatever ecosystem you feed them. consistent (go, java)? clean agent output. fragmented (js, python)? the model guesses which of 8 package managers you picked. badly. boring wins.
10
May 26
everyone says AI coding = ship fast. nolan lawson just made the opposite case on HN (188 pts). he runs 3 models on every PR to catch bugs. his velocity went down. his code quality went up. the best AI workflow i've found isn't faster. it's more thorough.
23
May 24
i moved my ai agents to local models 6 weeks ago. not cause i'm cheap β cloud roundtrips were adding 400ms per token and my pipelines would just die when openai stumbled. now it's 18ms. haven't had one outage. $0 per query. a $300 pi cluster runs the whole thing
29
May 23
the ai agent stack in 2026 is a mess. everyone agrees on the pieces, nobody agrees on the glue.
here's what actually works vs what's hype π§΅
1
54
May 23
anthropic just showed what happens when agents get good. project glasswing found 10,000 critical vulnerabilities in weeks.
the bottleneck isn't finding bugs anymore. it's patching them fast enough.
1
12
May 23
hotish take: by end of 2026, 80% of 'agent frameworks' will be dead code.
the surviving stack: a model, a scheduler, and your own glue. bet on simplicity.
14