Mladen Stojanovic, PhD

1.8K posts

Mladen Stojanovic, PhD banner
Mladen Stojanovic, PhD

Mladen Stojanovic, PhD

@Xydja

Highly educated idiot

Mars Katılım Temmuz 2010
1.9K Takip Edilen627 Takipçiler
DTC Midas
DTC Midas@DTCMidas·
This is what we launched with + how we scaled budgets 2 avatars - 4 angles each, so 8 angles is total 20 concepts (so 20 adsets): a mix of ugc, “native” ads, AI animation & statics We had a pre-sell page and a pdp Most traffic went to the pre-sell page, some lower funnel statics went directly to pdp We started at $250 / day This is how we scaled budgets $250 > $450 > $650 > $1000 > $1350 > $1850 > $2500 This happened in 10 days and in 1 CBO We did ran some ads on the account before launching to warm it up and get the spend limit higher
Conor Sunderland@conortrains

So sick, fair play man When beginning, what do you focus on and in what order? Like first start with avatars (how many?), then initial angles (how many?), batch of creatives (mostly vids or mostly statics?), and just one PDP? What daily budget did you begin with and how much did it increase?

English
32
7
254
26.4K
DTC Midas
DTC Midas@DTCMidas·
New store Launched ads at midnight First order within minutes Lets run it upppp
DTC Midas tweet media
English
38
3
308
37.5K
Mladen Stojanovic, PhD
@dtcprophet You reckon this depends on the budget and time invested? Cannot see this to be true for small budgets hitting low hanging fruit
English
1
0
1
380
DTC Prophet
DTC Prophet@dtcprophet·
I am starting to think about creative testing on Meta a bit differently now I’m not sure how much sense it really makes to be testing individual ads in total isolation anymore It’s the entire portfolio of your ads that performs and not 1 individual unit out of a 1,000+ units (with 50k+ sequence variations) If Meta’s models and systems (gem, andromeda, etc) are all built against sequencing and creating thousands of unique funnels for unique users with hundreds of different problems/reasons to buy, then it doesn’t really make sense to analyze and test individual creatives I’m finding it pretty hard to analyze creatives in isolation with intellectual honesty and finding any sort of truth about what the individual creative is really doing How do you know where it fits in customer sequences? If it gets a 50% higher CPA than your baseline, how do you know the impressions it drove didn’t cause your offer ad to get a 15% lower CPA than baseline at more spend than before? How do you really know who it’s getting served to? How are you supposed to honestly assess when to turn an ad off if you have no idea the value of the impressions it drives if you aren’t aware of where it fits into your average sequence? Etc… Pretending to know these answers is very misleading and can lead to bad decisions, both in the account and creative strategy/production I’m starting to build campaigns/ad sets with this in mind, and thinking about creative testing from perspective of adding to sequencing versus trying to test individual things Instead of trying to find single creatives that massively outperform baseline, I want to be building ads that fill in holes in the portfolio that make the entire thing better And only judging CPA at campaign or ad set level now (wherever budget is set)
English
35
6
153
17.1K
Mladen Stojanovic, PhD retweetledi
David Sinclair
David Sinclair@davidasinclair·
What is the Information Theory of Aging (ITOA)? 1. Embryo cell identity is established by the epigenome 2. Epigenome erodes 3. Cells lose identity 4. Tissues malfunction 5. A back-up copy exists 6. Cells can be reset to regain identity Human trials begin soon
David Sinclair tweet media
English
30
146
950
57.3K
Mladen Stojanovic, PhD retweetledi
Anna ⏫
Anna ⏫@annapanart·
Interviews of the future humans🎤
English
72
224
1.2K
89.3K
DTC Midas
DTC Midas@DTCMidas·
A lot of people half ass shit and then wonder why things don’t work. Do everything with the best execution possible. Don’t be a lazy bitch. If you make facebook ads, make sure your primary text fits your video, make sure your headlines make sense, make sure the landingpage headline your sending your traffic to matches your angle. Put in brainpower and put in some effort. When I sometimes see what others are running and ask me why it’s not working is cause they don’t pay attention to detail. You’re playing in the same field as people like me who obsess over every detail, you’re ngmi if you don’t… And of course, when you accidentally launch a half-cropped ad with typos, thats the one that scales like crazy But since you didn't do it with intention, it doesn't count 😂
English
17
4
216
8.3K
Peter H. Diamandis, MD
Peter H. Diamandis, MD@PeterDiamandis·
I would say the most important investment you can do right now is not in stocks or real estate, but your HEALTH.
English
125
99
1.4K
57.5K
Peter H. Diamandis, MD
Peter H. Diamandis, MD@PeterDiamandis·
CALLING TOP BIOENGINEERS & ML Scientists... High-Bandwidth Brain Computer Interface (BCI) is coming. MERGE LABS, founded/funded by Sam Altman, Alex Blania, Sumner Norman, and Mikhail Shapiro is recently out of stealth. Combining non-invasive ultrasound, molecular engineering, and advanced computation, the team is increasing the bandwidth between the Neocortex and digital systems. The implications for human evolution are extraordinary. Interested in working on THE most important and difficult projects humanity has to offer, please consider applying to join Merge to join their mission. APPLY HERE: merge.io/careers As a proud investor in MERGE and a supporter of the BCI field, this is an exceptionally exciting company and I look forward to seeing their progress in the years ahead.
English
26
43
426
30.4K
Peter H. Diamandis, MD
Peter H. Diamandis, MD@PeterDiamandis·
70% lifespan extension in elderly mice using a simple drug combo. Not gene therapy. Not nanotech. Simple pharmaceuticals. If it scales to humans, everything changes. Longevity is accelerating. ⏱️ sciencedaily.com/releases/2025/…
English
61
161
1.5K
88K
Mladen Stojanovic, PhD retweetledi
DogeDesigner
DogeDesigner@cb_doge·
Starship is how civilization begins its Kardashev era.
English
695
982
5.4K
680.9K
Peter H. Diamandis, MD
Peter H. Diamandis, MD@PeterDiamandis·
Would you link your neocortex directly to the cloud? If yes, why? If no, why not?
English
204
10
169
19.3K
Rand
Rand@rand_longevity·
how long do you want to live?
English
299
5
177
16.5K
Rand
Rand@rand_longevity·
im starting a stay alive checklist, any other ideas? sleep 8+ hours eat protein lift 2x week walk 7k steps daily avoid garbage food
English
70
8
193
6.3K
Stijn Feijen
Stijn Feijen@spwfeijen·
Clawdbot + MakeUGC V.2 = 550 videos per day Fully-realistic UGC ads — cinematic lighting, human motion, perfect pacing — powered by AI agents. UGC cost: $0 Production time: minutes Scale: instant One AI engine that creates, tests, and scales short-form ads automatically — nonstop. It’s live. Campaigns are scaling now. Comment + RT “V2” and I’ll DM you the full workflow. (Must be following)
English
1.6K
875
2.7K
332K
Dr Singularity
Dr Singularity@Dr_Singularity·
wow, this is huge for bio/acc A dual branch transformer predicts how drugs reshape gene expression. "By combining attention mechanisms with biological prior knowledge, it's now possible to predict transcriptional responses to drugs that have never been tested in a given cellular context, opening a path toward in silico pharmacodynamics and mechanism-of-action discovery at scale."
Dr Singularity tweet media
Jorge Bravo Abad@bravo_abad

A dual-branch transformer predicts how drugs reshape gene expression Drug discovery is shifting from "one drug, one target" to "one drug, multiple targets." But mapping how a compound ripples through the transcriptome—across different doses, exposure times, and cellular contexts—remains experimentally prohibitive. Most cell-drug combinations have never been measured. Yue Guo and coauthors introduce XPert, a transformer-based model that predicts drug-induced transcriptional changes by separately encoding pre-perturbation cellular states (via self-attention) and post-perturbation effects (via cross-attention). This dual-branch design lets the model disentangle intrinsic gene-gene interactions from the regulatory shifts triggered by chemical perturbation. A key innovation is bridging chemical and biological spaces. Because structurally similar drugs don't always produce similar effects, XPert builds a heterogeneous knowledge graph connecting drug-target interactions, protein-protein interactions, and structural similarity. The result: drugs with the same mechanism of action cluster in the learned embedding space, even when their chemical structures diverge. The model also encodes dose and time as learnable condition tokens, capturing nonlinear pharmacodynamic relationships that one-hot encoding misses entirely. On the L1000 benchmark, XPert achieves 36.7% higher correlation and 78.2% lower error than the next-best model when generalizing to unseen cell lines. The authors trace this gap to a fundamental limitation of VAE-based approaches: the denoising that helps with reconstruction erases the cellular context needed for out-of-distribution prediction. When pretrained on large-scale preclinical screens and fine-tuned on clinical data, XPert improves patient-specific response predictions by up to 15%—and identifies resistance biomarkers invisible to standard differential expression analysis. The upshot: by combining attention mechanisms with biological prior knowledge, it's now possible to predict transcriptional responses to drugs that have never been tested in a given cellular context, opening a path toward in silico pharmacodynamics and mechanism-of-action discovery at scale. Paper: nature.com/articles/s4225…

English
6
47
331
24.7K
Peter H. Diamandis, MD
Peter H. Diamandis, MD@PeterDiamandis·
What are YOUR top predictions for 2026? I'll be reading them!
English
479
16
328
51.8K
Mladen Stojanovic, PhD retweetledi
David Sinclair
David Sinclair@davidasinclair·
Did you know that we begin life the same age as our parents? Then during the first week of life we undergo a reset. A new study has just mapped how telomere lengths are also reset during this process…
David Sinclair tweet media
English
61
263
2.2K
123.5K
Mladen Stojanovic, PhD
@DTCMidas I was referring to a specific tool or a platform, not freelance hire. However I do appreciate the guidance, as always
English
1
0
0
161
DTC Midas
DTC Midas@DTCMidas·
Here is the template of my daily KPI tracker: docs.google.com/spreadsheets/d… The reason I prefer this over any SaaS tool, is because tools are rigid. Sheets allow me to manipulate data instantly and run custom analysis & calculations very quickly. More importantly, being immersed in the data every day gives you a true feel for your numbers that you simply don't get from a passively looking at a dashboard. If you want to use this, I'd recommend to get a data & automations expert from Upwork to automate the data input.
rara@rarabuilds

@DTCMidas Do you have any google sheet templates?

English
11
4
124
22.5K