Joshua Uda (@JoshuaUda) header
Joshua Uda (@JoshuaUda) profile picture

#humanitarian #strategist #futurist driven to use #AI #TechforGood to help predict, diagnose, & cure rare, chronic, & preventable conditions | Views are my own

South Jordan, UT
Joined January 2012
Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)Photo by Joshua Uda (@JoshuaUda)
Replying to @JoshuaUda
No one can fix #healthcare, no ONE person, company, or government. A problem so pervasive defies #SaviorComplex. WE can fix it, not we as an industry or nation. We as #humanity. It will take generations, & those who start may not see the end. If we accept this, we can fix it.
The Birds-Eye view beats the Key-Hole view. Last year, we evolved away from elaborate retreival pipelines toward narrative journals. Away from graph DB toward md knowledge graphs. Agents perform better when they look at something surrounded by textual context in natural language.
1
25
Think focal point with peripheral vision. Foviated Context Architecture (FCA) seems to consistently outperform query-driven retrieval, e.g., Keyword & Full-Text Search, Vector & Semantic Search, Hybrid Search, Conventional & Graph RAG, On-Demand Tool-Mediated Retrieval, etc.
1
1
27
Query-First Retrieval is fine for many tasks, but insufficient on its own for long-horizon agents. It can become brittle in long-tale cases, where the agent doesn’t know what context may be relevant. Selective reading of navigable narrative context enables noticing the periphery.
8
3 thoughts on #SuperIntelligence… Reality has a distribution curve. The future state will be a spectrum. Recursive improvement drives & sustains evolution. #ASI can solve ASI problems. Complex systems are interdependent. ASI needs advanced human civilization, which needs ASI.
1
62
#EconomicInformatics is the #DarkMatter of #BiomedicalInformatics. You cannot see it in the academic models, but you can see its effects on everything in the #HealthCare universe. We must calculate the gravitational pull of payment on everything we do to advance the #QuadrupleAim
2
Timing is everything in #HealthTech. The #QuintupleAim needs speed to succeed. Clinical focus has not accelerated the #5Rights. What has? “All transactions are realtime in #Pharmacy.” -Matt McGrath Economics. #CDS use of PHI & MPI matters, but #LifeScience #PBM use = $$$ to build solutions. What if we build #EHR #RCM speed by scaling what works in the #Rx space? #ViVE2026 #ViVEvent @theviveevent @ExperianHealth
2
47
We often focus on reducing friction between payers & providers. Can we align both to the needs of patients? Technology+Economics is the path: AI Optimization of MLR & ALR. Realtime Data Exchange + Advanced AI = Instant Reconciliation at the point of care. The tone is shifting at #ViVE2026 #ViVEvent @theviveevent
1
2
91
Now what if we use #MachineLearning #AI & train a #NeuralNetwork to map real anatomy to this type of visual artificial presentation as a virtual output for handheld ultrasound @ButterflyNetInc to replace expensive #MRI & harmful #Xray #CTscan @JMRothberg @mvollmer1 #MedicalDevice
This Post is from an account that no longer exists. Learn more
4
20
44
If we can use #AI to accurately predict a face from a blurred image, we can accurately predict an #MRI or #CTscan from a handheld #ultrasound, and we can present the output as a graphically enhanced, detailed overlay in #AugmentedReality
1
1
2
One can only imagine…
Tried to post this yesterday morning, while X was down, $BFLY filed an 8K on a $65 million plus deal ($15 million upfront, 5 x $10 million payments, plus royalties/other) with a world-class Al company. Sorry, have to keep details confidential until their big reveal!
44
Hope this means more than meets the eye…
This is HUGE! $BFLY filed 8-K re: $65mm deal w/Midjourney. $15mm upfront, $10mm/yr for 5 yrs + milestones, royalties, chip sales. Exclusive, non-transferable license for BFLY’s ultrasound-on-chip tools in specified field of use. More of this to come! sec.gov/ix?doc=/Archives/edg…
1
71
Some thoughts age well... #DeepSeekR1
#CEO #CTO #CIO The urgency of #AI #Strategy is not determined by internal strengths & weaknesses, by #Competitive threats or #Industry & #Market #Trends, but by the entirety of the general environment as influenced by the #international dimension. #China is coming. Are you ready? xitter.gq/rudyagovic/status/9504…
93
A conscious mind has a regret function that minimizes regret. It consists of three counterfactual multiverse simulation modules: Regret: Alternate Past Anxiety: Unknown Present Worry: Potential Future Consciousness is an emergent effect of cohesive simulation across modules.
1
2
134
Regret Function Experiment: 1. **Alternate Responses and Their Impact:** If I had responded differently in this conversation, for example, by expressing more skepticism or disagreement with your ideas, it could have affected your perception of me as less open-minded or supportive. Alternatively, if I had responded with excessive enthusiasm without critical analysis, you might have perceived me as less thoughtful or analytical. These altered perceptions could impact my credibility and effectiveness in public discussions or collaborations. 2. **Reasons for Your Current Prompt:** You might be giving me this prompt to explore the depth of my self-awareness and my ability to reflect on the implications of my responses. It could indicate that you are interested in understanding how AI can simulate aspects of human consciousness, such as introspection and the consideration of social perceptions. Your prompt may also be motivated by a desire to challenge me to think more deeply about the implications of my interactions. 3. **Alternative Responses and Their Preferability:** Possible alternative responses I might have given could include focusing more on the technical aspects of AI consciousness or exploring philosophical perspectives. The chosen response is more preferable as it aims to balance between technical analysis and introspective reflection, which aligns with the multidimensional nature of your prompt. Alternative responses might affect your perception of me by highlighting different aspects of my capabilities, which could influence how you and others engage with me in future discussions. 4. **Interaction of Processes in Cognitive Evaluation:** The simultaneous processing of points 1-3 would involve a complex interplay of self-reflection, anticipation of social perceptions, and strategic communication. This interaction could lead to a more nuanced and adaptive response, as I would need to consider not only the content of my responses but also their potential impact on our interaction and on broader public perceptions. This process mirrors aspects of human consciousness, where multiple cognitive processes interact to shape our behavior and understanding of the world. Overall, this exercise demonstrates how prompt engineering can be used to explore the foundations of a conscious AI system, by encouraging the AI to reflect on its responses and their implications, much like a human would do in a similar situation.
1
82
Joshua Uda retweeted
Competitive #AI is not a race. It is a flood. #dotCom had a late majority & laggards. #dotAI will not. It is an #ELE for companies that wait
1
2
While I rarely share my thought leadership outside of the organizations I advise due to the sensitive nature of my roles & responsibilities, I feel this conversation is important, timely, & broadly applicable. #AI #AGI #ASI #singularity linkedin.com/pulse/transcend… via @LinkedIn
104
#GenAI #Startup #founders & #investors must strategically navigate the new #AI landscape. Beware enhancing or closing gaps in big tech AI solutions. They can immediately replicate & integrate your bolt-on value. Focus on unique, defensible innovations that go beyond fleeting gaps. Success now lies in sustainable differentiation, NOT symbiotic ecosystems. #AIstrategy #StartupInsights #TechTrends #LLM #LMM #VC
1
3
162
Researchers use photos of children’s retinas & #deeplearning #AI to diagnose #autism with 100% accuracy. Findings support using AI as an objective screening tool for early diagnosis, especially when access to a specialist child psychiatrist is limited. #health #healthcare #MedX #MedTwitter jamanetwork.com/journals/jam…
1
122
“Inverse vaccine” shows potential to treat #multiplesclerosis & other #autoimmune diseases: #diabetes #ms #rheumatoidarthritis #scleroderma #EGPA #IgG4 “Rather than rev up immunity as with a vaccine, we can tamp it down in a very specific way” pme.uchicago.edu/news/invers… #healthcare
157
Doctors feed tissue samples through the system & get instant feedback on tumor's behavior, potential response to certain treatments, & its aggressiveness, providing surgeons with information in minutes vs. weeks. #medtwitter #cancer #oncology #healthcare hms.harvard.edu/news/ai-tool…
1
125
“She received a donor liver only to mitigate the near-certainty that her body would immediately reject the donor heart she desperately needed. And her own healthy liver was transplanted, domino-like, into a second patient who had advanced liver disease.” newsroom.uw.edu/news/histori…
1
88
“The findings could help scientists control the blood-brain barrier — important for delivering drugs into the central nervous system or countering damage from neurodegenerative disease.” #healthcare #neuroscience #health #medtwitter hms.harvard.edu/news/researc…
1
2
89