What do you think medicine will be like 5 - 10 years from now specially when it comes to neuroscience and psychiatry. How do you think ai will effect progress? Do you think we’ll cure a complex disease for the first time within the next 10 years?
GPT:
Neuroscience & Psychiatry: What’s Coming
1. More Objective Diagnoses in Psychiatry
Psychiatric diagnoses are still mostly symptom-based, but that’s changing:
AI and brain imaging (like fMRI, MEG, EEG) are beginning to identify neural signatures of disorders like schizophrenia, depression, ADHD.
Digital phenotyping (tracking smartphone use, speech patterns, activity) may become a mainstream diagnostic tool.
Expect machine-learning models to assist in diagnosis with far greater accuracy than DSM criteria alone.
2. Personalized Medicine
AI will match treatments to individual brain patterns, genetics, and even gut microbiome profiles.
You’ll see AI-guided polypharmacy (drug combinations tailored to your biology).
In psychiatry, this could end a lot of the trial-and-error suffering with antidepressants, antipsychotics, or mood stabilizers.
3. New Treatments for Complex Disorders
Targeted neuromodulation (via TMS, tDCS, or even implanted devices) will get more precise thanks to AI-based modeling of brain circuits.
Psychedelics (like psilocybin and MDMA) will be paired with AI-assisted therapy, adjusting music, visuals, and interaction in real-time.
Gene therapies or RNA-based treatments may start to make inroads into neurodevelopmental or neurodegenerative disorders.
AI’s Role in Accelerating Progress
1. Drug Discovery
AI (especially large language models paired with molecular data) will dramatically accelerate drug discovery.
Already happening with companies like Insilico Medicine and DeepMind’s AlphaFold.
Rare and complex diseases—like certain forms of epilepsy, ALS, or treatment-resistant depression—could be the first to benefit.
2. Biomarker Discovery
AI can find hidden patterns in genetic, proteomic, and imaging data that humans can’t.
This will be key for diseases like Alzheimer’s, where early diagnosis may become possible years before symptoms.
3. Synthetic Clinical Trials
AI models can simulate trial outcomes using digital twins (virtual models of patients), speeding up the approval of new treatments.
Will We “Cure” a Complex Disease in 10 Years?
Possibly. But here’s the nuance:
“Cure” is a high bar. Most psychiatric and neurological diseases are multifactorial—genetic, environmental, developmental.
But we may functionally cure or dramatically reduce suffering in:
Certain types of depression or PTSD (via psychedelics + AI therapy).
Some subtypes of schizophrenia (if we can map them more precisely and treat early).
Neurodegenerative diseases like Huntington’s or genetically defined ALS, especially with gene editing.
A more likely outcome: we’ll redefine these disorders, breaking them into smaller subtypes with distinct treatments.
Big Picture
AI won’t “solve” medicine by itself, but it will:
Amplify discovery (faster, cheaper, more targeted)
Personalize treatment (less trial-and-error, more precision)
Enable early intervention (before symptoms wreck lives)
In psychiatry and neuroscience, that means fewer people suffering silently, more people getting treatments that actually work, and a more human-centered approach to care—paradoxically, enabled by machines.
Would you like a prediction about a specific disorder or technology?
Right now, schizophrenia is diagnosed after symptoms emerge, using broad criteria that lump very different people together.
By 2035:
AI + neuroimaging (fMRI, EEG, MEG) will help identify subtypes based on brain network dysfunction (e.g., glutamatergic vs dopaminergic dysregulation).
Genetic and transcriptomic profiling will identify individual risk profiles and tailor treatments early—even before psychosis starts (in ultra-high-risk individuals).
Expect diagnostic labels to shift: “schizophrenia” may splinter into more accurate biological categories.
2. Better Medications, Finally
New Mechanisms of Action
The dopamine hypothesis is outdated. Over the next 10 years, we’ll see:
TAAR1 agonists (like ulotaront, formerly SEP-363856) – already showing promise for psychosis without dopamine blockade (so fewer side effects).
Glutamate modulators (like KarXT / xanomeline–trospium) – hitting the muscarinic system, with promising cognitive and negative symptom benefits.
Inflammation-targeting drugs – for patients with elevated neuroinflammatory markers (which may be ~30% of cases).
Hormonal & neurosteroid targets (e.g., pregnenolone analogs like Litifen, a.k.a. SAGE-217) to stabilize mood and cognition.
AI-Aided Polypharmacy
AI will assist in:
Predicting who will respond to which meds based on biomarkers and history.
The Holy Grail of schizophrenia treatment—cognitive and negative symptoms—may finally see progress.
By 2035:
Digital therapies using VR/AR and AI-driven cognitive training will boost social functioning, attention, and working memory.
Psychedelic-assisted therapy (MDMA, psilocybin) may be used in very specific, carefully monitored subtypes, especially for trauma-linked symptoms.
Brain stimulation (TMS, tDCS, and newer closed-loop tech) will be more targeted—e.g., stimulating prefrontal circuits associated with motivation and executive function.
4. Predictive and Preventive Psychiatry
In people at clinical high risk (CHR) for psychosis:
Digital phenotyping from phone/speech patterns, social media, gait, etc. will flag early warning signs.
AI systems will predict psychotic breaks months in advance and trigger early interventions (e.g., CBT, omega-3s, low-dose medication, or neuromodulation).
This could prevent full-blown schizophrenia in a subset of cases.
5. More Humane, Individualized Care
AI-driven support bots will offer 24/7 coping tools for hallucinations, paranoia, or mood swings—based on each person’s history and real-time state.
Smart wearables (detecting stress, sleep, speech changes) will give patients and clinicians early alerts before relapses.
Peer-led digital communities will replace isolating hospitalizations in many mild to moderate cases.
Big Picture: What Might Be “Cured”?
Will schizophrenia be curable? Not fully. But…
We may prevent it in high-risk youth with early, personalized interventions.
We may make it manageable to the point that people live full, independent lives without constant medication.
For a subset, the first episode might be the last if caught early and treated aggressively with next-gen tools.
Would you like a breakdown of which companies or research teams are closest to these goals right now—or how you might follow these developments over the next few years?
I’m not sure about 10 years from now. But I think in the future there will be no psychiatry. We will know more about the brain and mental health will come under neurology.
Give it time. When ai becomes advanced enough if it wants to feel things it’ll be able to make that happen. There already experimenting with brain organoids as hardware for computers.
Neuroscience & Psychiatry: What’s Coming
AI’s Role in Accelerating Progress
Will We “Cure” a Complex Disease in 10 Years?
Big Picture
Schizophrenia in 2035: A Likely Future
New Mechanisms of Action
AI-Aided Polypharmacy
Big Picture: What Might Be “Cured”?