10 Sep 2025
Imagine training a basketball team by only watching highlight reels. They’ll look flashy but collapse in a real game.
The big picture = AI can’t be truly intelligent until its data reflects all of us, in all our complexity…
#frontierdata #DGN #dataengines
10 Sep 2025
🚨My most serious tweet ever (can we even call it a tweet anymore? My most serious ‘X’ ever? She sounds terrible…)
@DataGuardiansNK is solving the Fei-Fei Li problem. That’s why $DGN will be a mega cap token!
Current AI models mimic language. They fail when asked to reason about the physical or social world. At D-GN we are building the provable governance layer for frontier datasets that embed embodied common sense, the missing piece for AI to cross from text imitation into true reasoning. Without YOU there is no #AGI
🚀🌕
#AI #DGN #PumpFun
3
8
13
346
24 Jan 2024
Thank you @dataengines for providing this additional paper. I've skimmed through it and will definitely read it.🙏
1
2
26
But language models are not authorities. There are other places to look for more authoritative responses.
3
2
29
27 Oct 2023
I agree with this point: I found the pars construens of the paper (6. Alternatives to ethical principles) much weaker than the pars destruens.
2
1
13
17 Apr 2023
📢 don’t forget to join us TODAY @ 5:30 pm in Starr Foundation Hall at 📍63 5th avenue
@yunnia from @MichiganESC will discuss #dataEngines The Allure of China’s Soil and Soul
Co-organized with @NewSchool_IA
RSVP: event.newschool.edu/chinastr…
2
192
25 Jan 2023
Is the Transition to Service-oriented Software Engineering Already Underway? bit.ly/3XTucyb #dataengineering #dataengineeringservices #dataengineer #dataengineeringsolutions @dataengines @DataEngineerPro @Mr_DataEngineer
1
66
16 Jan 2023
#AI adoption: is it obvious yet?🧐
People in established organizations don’t naturally think of AI when framing how they will solve problems; they therefore don’t build #dataengines, which is how competitive advantage is nurtured in AI space. @neal_lathia👉nlathia.github.io/2023/01/Is…
3
292
4 Jan 2023
In mediation analysis, we can compute the direct and indirect effects, we even can partition total effects into the natural direct and indirect effects but that doesn't mean these effects are actually present in the changes occurring in a treated unit.
2
66
4 Jan 2023
Dr. Cartwright argued that just because we can analyze something in our heads does not mean we can reduce biological causality to physics.
As to ATE not having anything to do with the unit-level causal effect, it’s established fact: PNS ≠ ATE in general
2
65
4 Jan 2023
Well, then you shouldn't have any trouble sticking your hand in the bucket of water and pulling out a handful of oxygen. 😂
1
2
62
1 Jan 2023
This is a very clever and novel usage of the term ‘Laplace’s Demon’. I like it a lot! The fundamental non-computability of biological systems may mean there is no shorter way of modeling their behavior, than to ‘run’ their programs. It is also a barrier to synthetic redesign.
2
41
23 Dec 2022
@Henry_Wynn is, of course, a leading figure in the theory of experimental design.
2
2
154
22 Dec 2022
For the record, I use the "demon" to reference Laplace's demon, who was supposed to predict any future event from the position vel of all particles in the universe. Instead, knowing the (genetic) cell-level context fails to predict organism-level properties.
2
10
699
Computational Fluid Dynamics just published nextcfd.com Thank you @dataengines @dr_umitgunes @GnosisGuild #bioinformatics #computational
2
3
1,439
21 Dec 2022
These are really interesting/deep points... all of them falling way out of my field of expertise! But if we take this as a question of (good-enough) prior belief formation without infinite regression, isn't that what a NESS density could offer us as embodied intelligent agents?
1
2
57
21 Dec 2022
Thanks for prompting some interesting thoughts this morning, and for following me also! 🎉
1
2
44
21 Dec 2022
Sure. One cannot escape our situation as observers and modellers to some absolute ground truth. But there is a formal canonicity to free energy principle theoretic models of belief updating that IMO helps us avoid reification or regress, as we argue here: arxiv.org/abs/2208.06924
1
1
60
21 Dec 2022
Fair question. I’d point to the multi-scaleness of the free energy principle formulation: hyperparameters of densities at one scale are set by dynamics at superordinate scales, see e.g.: link.springer.com/article/10… or sciencedirect.com/science/ar…
2
3
119
19 Dec 2022
Perhaps one way to start is that @yudapearl understand (finally!) that experiments does not require representative samples (this is a very old idea in bioestatistics in fact).
Even more, many times is not possible to reach generalizable conclusions with this type of samples!
2
2
188