Matt Farnell
1.2K posts
Matt Farnell
@farnell
📍software & sabbaticals
SF & Sydney Katılım Haziran 2008
1.4K Takip Edilen849 Takipçiler
@creatine_cycle the context window is incredible too
she remembers something you said 3 years ago in a completely different conversation and brings it up at the worst possible moment
no rag needed)
English
Matt Farnell retweetledi
we’re about to speedrun this realization across every institution in society aren’t we
staysaasy@staysaasy
I’m on week five of trying to vibe code a replacement for some dumb saas that we use and it’s so incredibly frustrating that I’m slowly realizing it’s actually a quite complex and thoughtful piece of software.
English
@Arash_Meh @CollinRugg mate, relax and enjoy your life a little bit
English
@CollinRugg No real work can be done like that… just a fancy adult day care
English
I just personalized 1 million emails with Claude Code.
I didn’t use any enrichment platforms. Here’s my 14-step workflow:
1. Install Python 3.11+
2. Create a GitHub account
3. Install Claude Code
4. Create API keys:
→ Anthropic (Claude)
→ Proxy (IPRoyal or Bright Data)
→ Company data (Clearbit, Apollo, Crunchbase)
→ Email verification (NeverBounce or Reoon)
→ Storage (AWS S3 or GCS)
→ Database (Neon, Supabase, Railway)
→ Outreach tool (Instantly, Smartlead, etc.)
→ CRM (Attio, HubSpot, etc.)
5. Create a new GitHub repo (“Enrichment”) and clone locally
6. Open terminal in the repo and launch Claude Code
7. Tell Claude your constraints:
→ Respect robots.txt
→ Rate limit requests
→ Make it resumable
8. Paste this prompt:
“Build a Python project that:
– Takes input.csv with linkedin_url, website_url, or company_name
– Fetches website, careers, blog, pricing
– Extracts: hiring signals, pain signals, and 1-sentence personalization hook
– Outputs output.csv
– Supports concurrency, retries, logging, resume-from-last-row”
9. Test locally:
pip install -r requirements.txt
python main.py --input input.csv --limit 10
10. Review output.csv and confirm hooks look good
11. Tell Claude to productionize it:
→ Env vars
→ Backoff + retries
→ Caching
→ Timeouts
→ do_not_contact flag
→ Unit tests
12. Deploy on Railway → connect repo → add API keys → deploy worker
13. Upload lead CSV → process in chunks → download enriched output.csv
14. Push enriched leads directly into your outreach tool
You now have fully personalized leads at scale. No enrichment platforms required.
Found it helpful? I run a GTM community where we share free playbooks and workflows like this weekly.
Reply “JOIN” and I’ll send you an invite.
English
Matt Farnell retweetledi
AGI is not coming.
We are nowhere near AGI. What we have today is inference, not learning.
Models get trained once on huge fixed datasets, then frozen. You ask questions, they remix patterns they already saw. Nothing updates. Nothing sticks. Talking to the model does not make it smarter. It does not learn from you. Ever.
Learning is still slow, expensive - and offline.
Look at self driving. You drive around a pothole, make a U turn, and come back. The car’s AI does not learn that you just solved that exact problem. It reacts the same way every time using sensors and rules. Do this 20 times a day and it still has zero memory that the pothole exists. It just re sees it. That is why edge cases never die. There is no local learning. No accumulation.
No 'oh yeah, I’ve seen this before'
LLMs work the same way. Tell it your name and it does not remember. The only reason it looks like memory is because scaffolding keeps shoving your name back into the prompt every time and sanitizing the output.
The model itself has no idea who you are and cannot learn from interaction. It is structurally incapable.
And the scaffolding is the worst part. It is pure duct tape. Just prompts on prompts on prompts around a frozen model. When something breaks, nobody fixes learning. They add another layer. Another rule. Another retry. Another evaluator model judging the first model.
So you end up with systems that are insanely complex but mentally shallow. Debugging is hell because behavior comes from hack interactions, not a learnable core. Tiny prompt tweaks cause wild behavior shifts. Latency goes up. Costs go up. Reliability goes down. None of this compounds into intelligence. It just hides the cracks.
Until we have real persistent learning and real memory inside the system, there is no AGI.
LLMs are not built for this. You cannot prompt your way out of it. You need a totally different architecture. Yann LeCun is right.
And even then, what architecture can actually learn online, store memory, and stay stable on today’s hardware?
Best case, maybe 5-10 yrs.
Right now it is all inference. It looks magical, but the emperor has no clothes. A lot of people see it. Almost nobody says it out loud.
English
Optimus will be amazing for protecting and taking care of your elderly parents
Adam Lowisz X Meetup 🇺🇸🇵🇱🇪🇺🇬🇧🇺🇦@AdamLowisz
This is a great discussion by @elonmusk at the WEF in Davos on how Optimus will help the elderly. Our population is aging rapidly. Just the other day my grandmother fell and broke her hip. She lives by herself about an hour drive from me, so it's difficult to be there all the time. She was found on the floor laying for 5 hours. The only way we knew something was wrong was because she wasn't answering the phone, we had to drive an hour to see what happened. She refuses a caregiver who lives in the neighborhood to come in and check in on her, she insists that she doesn't need help. Optimus would be a huge lifesaver for her. She can still maintain her independence and have Optimus there to help her when she needs it. Optimus can adjust its help based on her personality and observing what she needs help with so she can still feel independent and active without all the dangers that come with being nearly 90 years old.
English
R.I.P lead gen agencies.
I just replaced an entire lead gen team with Claude agents.
(all working while I slept)
Most founders spend $10k-$20k/month on marketing teams that work 9-5.
Most agencies spend $30k+/mo on outreach.
Last night I built AI agents that run 24/7:
- Lead Magnet Engineer → builds viral lead magnets in minutes
- Social Media Expert → writes scroll-stopping hooks
- Creative Director → generates on-brand visuals
- Research Analyst → finds trending topics in your niche
- Performance Tracker → analyses and maps out content
The results after 24 hours:
- 32 lead magnets ready to launch
- 60 days of content mapped out
- 50+ scroll-stopping visuals created
While I was sleeping.
Follow + reply CLAUDE and I’ll send the full system + setup.
English
context window issues - solved (mostly) 🚀
Robert Youssef@rryssf_
DeepMind just did the unthinkable. They built an AI that doesn't need RAG and it has perfect memory of everything it's ever read. It's called Recursive Language Models, and it might mark the death of traditional context windows forever. Here's how it works (and why it matters way more than it sounds) ↓
English
Creating a PRIVATE Telegram group for people to learn about AI and vibe coding.
We’ll talk about what we’re building, revenue, favorite AI tools, plugins, Claude Code, new releases, marketing, content, X, etc.
Comment ‘AI’ if you want to join. I'll DM you the link.
English
I want to start a community dedicated to Claude Code.
It’s become the gateway drug to coding and experiencing the power of AI for tons of people.
This will be a space for people to share killer use cases, agentic workflows, proven prompts, and connect with other CC obsessives.
Comment “Claude” if you want to join.
English
Man, just the absolute state of Canadian healthcare. People die in the waiting rooms because there’s not enough doctors and the lines are too long.
RTN@RTNToronto
44 year-old man passes away in the hospital after waiting over 8 hours in the emergency room in Canadian hospital 😳💔
English
💯 the deeper i get into AI the more this becomes apparent
dharmesh@dharmesh
Agreed. Here's the advice I give my son (who is 14): Some of the most valuable skills are systems thinking, functional decomposition (being able to tackle large problems by breaking them down) and building instinct for how to abstract away complexity for others. This is not going to change. Those things will be even more important in the age of AI.
English
Big upgrade to vibe coding in @GoogleAIStudio lands in Jan, but if you want to test early… 👇🏻
English
@Prachi_2050 Commenting on stupid engagement farming questions like this.
English
Matt Farnell retweetledi
yep, its happening the world over..
Chamath Palihapitiya@chamath
Here is Peter Thiel’s email to Zuck and Andreessen in Jan-2020 predicting socialism. Tl;dr too much student debt and lack of affordable housing keeps young people with negative capital for too long. And without a stake in the capitalist system, they will turn against it.
English