Co-founder @tortastudios Ex-Facebook, Ex-Microsoft. I build systems that make AI Growth workflows observable, repeatable, and scalable.
NC
Joined February 2012
- Tweets25.2K
- Following2.3K
- Followers13.7K
- Likes43.2K
In a few years, when we get a legit time machine, I'm going back to tell little me:
"you're going to watch computers pull data from other dimensions and computers conjure real worlds in real time, just a few years apart from each other."
And He's not going to believe a word of it.
Explainer videos are neither easy nor fast.
anyone who tells you "just have claude do it" it's trying to sell you something.
I think we're getting better at creating them fast, but there's very little GenAI in them, and letting claude do the animations unsupervised, produces extremely low quality stuff.
proofingroom.tortastudios.co…
I should reverse engineer this, patch it and some dude getting pinged that “hey this shit that wasn’t working it’s actually working now. What do I do with this”
Ok, I just checked this, and its literally running a freaking TensorFlow Lite model on 960×540 images with INT8 quantization every couple of seconds in an attempt to figure out what game you are playing.
It's literally called "alex.tflite" and its literally AlexNet and seems to follow torchvision.models.alexnet. As far as I can tell this does some strange nearest neighbor feature comparisons against a DB and literally doesn't work 99% of the time, you love to see your CPU get wasted every other second by this thing though. It also just seems to dump its results into the void because the consent gate comes later.
Anyways, here's the tf model graph and some dumped captures.
The vulnerabilities were found in human-written, extensively human-reviewed code that had been in production for years before LLM generated code was a thing
Alan retweeted
Immigration policy must be driven by the present, not the past, and business leaders are too terrified of the left AND the right to acknowledge this.
Recent H1B discourse revolves around a longstanding immigration sleight-of-hand. H1B proponents point to exceptional people who came here decades ago as if they represent the current state of the program in the same way that advocates for low-skill/illegal immigration point to very favorable immigrant crime/welfare/education/etc statistics that do, in truth, have them beating native-born Americans on many fronts.
It is all a sham. The world that created those statistics, people, and businesses no longer exists. Brilliant but desperate Vietnamese fathers who fled communism have nothing in common with young men from Honduras, and the crime rates/economic acccomplishments/integration of the former should not be used to justify mass migration of the latter.
Nor should the accomplishments of people who managed to get an H1B visa in the 90s (when the program was capped to just 65,000 visas) be used to justify staffing agencies and tech companies using the program to replace American jobs for less money. The obvious abuse going on is politically suicidal and something that even the most ardent supporters of the visa should want to crack down on, no immigration program can possibly survive the obvious and widely known reality of what is going on. I am a longtime outspoken advocate for stealing the best people from other countries to make our own better, runaway abuse of existing programs robs the government of the worker trust required to do so.
Lots of people in the tech space know this. Even the dumb people are smart enough to see it. Nobody actually believes these are all incredible people with unique talents performing jobs no American can. Everyone knows the guy in Cubicle Bay 7 doing mediocre but functional web design for half the salary of his American peers is no savant and doesn't meet the statutory requirements for his H-1B visa.
Unfortunately, most people with power or profile are terrified that any public statement or acknowledgement will get them attacked by the right (You guys have been screwing over American workers for years with visa abuse, you damn hypocrite!) or the left (You are racist, sexist, misogynist, and sucking up to this fascist administration!). Or, more likely, both. I know too many tech execs in particular who feel trapped, like the only winning move is to stay silent.
JD is blowing the whole thing wide open by putting facts and stats front and center that nobody wants to engage with. Critics will talk about anything and everything else but these damning facts and this obvious truth, that H1B visas serve neither their original purpose, legal purpose, invented purpose, or the American people.
Vice President JD Vance breaks down the data:
"If you are working as an H-1B visa in the United States of America, you’re earning $20,000 less than an American citizen hired in the same position. If you were brought in by one of these foreign outsourcing firms, you’re making $48,000 less than an American citizen in a similar position. So I’m sure that everybody here understands the game that is being played. You bring in indentured servants from outside the country, you lay off American workers and, if you’re a corporation, you make a ton of money by under cutting the wages of American workers, replacing them with people who frankly shouldn’t be in the United States of America to begin with."
My heuristica es la opuesta 😅
Si no está al tanto de todo lo que de habla en X, automáticamente asumo que está desactualizado
Quizá suene exagerado, pero los que están súper AI-pilled, si se acerca mucho al concepto de religiones. El comportamiento es muy similar.
Alan retweeted
Who’s going to tell the old school coders that there’s going to be nothing but vibe coding in like 6 months. All companies will need are people who can describe what they want properly, and someone to look it over. These tweets that make vibe coding seem stupid, are naive.
Who's gonna tell vibe coders about...
1. SQL Injection
2. XSS
3. CSRF
4. CORS
5. SSRF
6. IDOR
7. Path Traversal
8. Mass Assignment
9. Clickjacking
10. Content Security Policy
11. Security Headers
12. Input Validation
13. Server-Side Authorization
14. Row-Level Security
15. Least Privilege
16. RBAC
17. Secrets Management
18. .env in .gitignore
19. API Key Rotation
20. Secret Scanning
21. Hashing vs Encryption
22. Password Salting
23. bcrypt & Argon2
24. JWT Expiry
25. Refresh Tokens
26. OAuth 2.0
27. OpenID Connect
28. Session Fixation
29. MFA
30. Brute Force Protection
31. Account Enumeration
32. Timing Attacks
33. File Upload Validation
34. Webhook Signature Verification
35. TLS Certificate Expiry
36. HSTS
37. Encryption at Rest
38. Dependency Vulnerabilities
39. Supply Chain Attacks
40. Threat Modeling
41. Indexes
42. EXPLAIN ANALYZE
43. N+1 Queries
44. Full Table Scans
45. Connection Pooling
46. Database Migrations
47. Down Migrations
48. Transactions
49. ACID
50. Isolation Levels
51. Deadlocks
52. Optimistic Locking
53. Pessimistic Locking
54. Normalization
55. Denormalization
56. Foreign Keys
57. Unique Constraints
58. Cascading Deletes
59. Soft Deletes
60. UUIDs vs Auto-Increment
61. Cursor-Based Pagination
62. Full-Text Search
63. Read Replicas
64. Replication Lag
65. Sharding
66. Partitioning
67. Hot Partitions
68. Postgres VACUUM
69. Materialized Views
70. Backups
71. Testing Your Restores
72. Point-in-Time Recovery
73. Change Data Capture
74. Outbox Pattern
75. ORM Lazy Loading
76. CAP Theorem
77. Eventual Consistency
78. Strong Consistency
79. Raft Consensus
80. Leader Election
81. Split Brain
82. Quorums
83. Clock Skew
84. Distributed Locks
85. Consistent Hashing
86. Bloom Filters
87. Two-Phase Commit
88. CQRS
89. Event Sourcing
90. Race Conditions
91. Backpressure
92. Load Shedding
93. Bulkheads
94. Jitter
95. Retry Storms
96. Cascading Failures
97. Thundering Herd
98. Cache Invalidation
99. Cache Hit Ratio
100. TTLs
101. Cache Warming
102. Write-Through vs Write-Back
103. Redis Eviction Policies
104. CDNs
105. Cache-Control Headers
106. ETags
107. Dead Letter Queues
108. Poison Messages
109. At-Least-Once Delivery
110. Exactly-Once Semantics
111. Message Ordering
112. Kafka Consumer Groups
113. Partition Keys
114. Background Jobs
115. Webhook Retries
116. DNS
117. DNS Propagation
118. HTTP Status Codes
119. HTTP Method Semantics
120. Keep-Alive
121. HTTP/2
122. WebSockets
123. Server-Sent Events
124. Long Polling
125. gRPC
126. Protocol Buffers
127. API Versioning
128. OpenAPI Specs
129. Backward Compatibility
130. Request Size Limits
131. Compression
132. Latency vs Throughput
133. Egress Fees
134. VPCs
135. Subnets
136. NAT Gateways
137. Firewalls
138. Security Groups
139. Service Mesh
140. Ingress Controllers
141. Health Checks
142. Liveness vs Readiness Probes
143. Graceful Shutdown
144. Connection Draining
145. Stateless Services
146. Failover
147. Multi-Region
148. Availability Zones
149. Single Points of Failure
150. Disaster Recovery
151. RTO & RPO
152. SLAs
153. SLOs
154. Error Budgets
155. Incident Response
156. On-Call
157. Runbooks
158. Postmortems
159. Chaos Engineering
160. Graceful Degradation
161. Autoscaling
162. Horizontal vs Vertical Scaling
163. Cold Starts
164. Capacity Planning
165. Load Testing
166. Memory Leaks
167. OOM Killer
168. Structured Logging
169. Log Levels
170. Log Rotation
171. Distributed Tracing
172. OpenTelemetry
173. Correlation IDs
174. Metrics Cardinality
175. p99 Latency
176. Alert Fatigue
177. Error Tracking
178. Profiling
179. Flame Graphs
180. Infrastructure as Code
181. Terraform State
182. GitOps
183. Blue-Green Deployments
184. Canary Releases
185. Rollbacks
186. Feature Flags
187. Environment Parity
188. Staging Environments
189. 12-Factor Apps
190. Multi-Stage Builds
191. Non-Root Containers
192. Image Scanning
193. Resource Requests & Limits
194. Pod Disruption Budgets
195. Semantic Versioning
196. Lockfiles
197. Reproducible Builds
198. Branch Protection
Si le enseñan el video de harambe a sus hijos antes de ir al zoológico y les dicen que avientan a los niños que no se portan bien con los gorilas van a tener un lindo paseo familiar 👍
Alan retweeted
Things population genetics have proven in recent years:
- Out of Africa theory was wrong
- "West Eurasians" (White people) were selected for intelligence (Akbari et al., 2026)
- Sub-Saharan Africans have a signficiant percentage of "ghost DNA" belonging to a now extinct hominid species (Durvasula & Sankararaman, 2020)
- The ancestors of modern Northern Europeans invaded India and established Vedic civilization (Narasimhan et al., 2019)
The supposedly debunked "race science" of the early 20th C. was far more correct than post-war anthropology. Normie discourse is decades behind.
Dr. Eleanor Scerri of Max Planck Institute says the idea that humans originated in Southern Africa based on the fact that the oldest haplogroups are located there is unlikely. gea.mpg.de/person/52339/2944
Yo salí de un cbtis 🙋♂️
Hay de todo, y si hay mucho bandido y malandro, pero también hay un chingo de gente trabajadora.
La mayoría de mi salón trabajaba, estudiaba y apoyaba en su casa. A los 16-17 muchos ya nos habíamos comprado nuestro propios coches, si, ya casi para la basura, pero eran nuestros.
Al final del día eso forja carácter y acompañado con oportunidades y un poquito de inteligencia/visión puedes llegar bien lejos
Alan retweeted
first time trying out jev for industry classification
client handed me 180k websites with no industries. paying per record in clay or open ai/etc to enrich all of that gets expensive fast, so i'm testing jev as the cheap classifier.
the flow:
- python loads the homepage, pulls title, description, nav, page text + tech stack signals
- if the homepage fails (cloudflare, dead site), fall back to serper google results
- jev returns industry + confidence, ecommerce y/n, dynamic site, parked or dead
- csv, then into clay
starting with 10k to check the failure rate before running the rest. the real goal is figuring out the shape of their account list and which verticals to actually target.
Hice este video sin abrir Blender, un editor de video ni una herramienta de modelado 3D. Lo dirigí desde una terminal, escribiendo prompts y viendo previews.
Llegar a este punto fue menos automático de lo que lo hago sonar originalmente. Hubo caras que se estiraban al caminar, trabajadores con herramientas mal agarradas y música tan fuerte que tapaba la narración.
Cada preview terminaba en notas sobre encuadre, movimiento, iluminación o sonido. Y cuando un plano ya estaba aprobado, también tuve que insistir en dejar de tocarlo porque cuando queria re-ajustar la escena 6, mis agentes insistian en renderizar todo otra vez.
En el proceso, he desarrollado un montón de skills para organizar las ideas, realizar las ediciones, seleccionar efectos de sonido y ajustar el volumen y los picos optimizados para social media. Tambien un monton de sub-agentes para escribir titulos, descripciones, para generar thumbnails, y que el titulo, la descripcion, el thumbnail, el video y la narrativa todo trabaje junto.
Está lejos de ser edición de clase mundial. Pero con cada iteración me acerco un poco más a resultados profesionales de expertos.
El experimento anterior, de 51s consiguió 1.6K vistas y 6 suscriptores nuevos en menos de 24 horas. El 64.7% decidió quedarse a verlo en vez de saltarlo, y el tiempo promedio de reproducción fue de 38 segundos.
Son resultados iniciales, pero me dan ganas de seguir iterando. Mi hipótesis es que este nuevo video va a retener mejor y ayudarnos a llegar a más gente.
Veremos qué dicen los números.
Vamos a seguir publicándolos en YouTube y contando el proceso, incluidos los errores, en nuestro newsletter.
Dejo ambos enlaces abajo.