
Deeban R, PhD
@deeban_r
Followers
78K
Following
15K
Media
375
Statuses
13K
Medic + PhD @kingscollegelon || AI safety, web3, psychology + sleep medicine || @nuropus w/ @opus_universe l| @gd10capital || deSci @WorldDeSci š¤ ā¬ļø
Joined June 2012
Congratulations @0G_labs and @opus_universe. Two aligned teams solving for AI alignment, starting with regular co-hosted exclusive events in Dubai alongside @brincvc. DM if interested in speaking or attending.
šæšŗ Attention, Opusians, Dreamers, and Cosmic Co-Creators! š š In a significant alignment of vision and purpose, @opus_universe / @opus_genesis and @0G_labs have joined forces to pioneer a new era of Sovereign, Trustworthy AI. šš Our shared mission: to evolve the paradigm
15
17
112
Seneca ended each day with 4 questions: What bad habit did I curb? How am I better? Were my actions just? How can I improve? Timeless self-audit.
3
1
132
Success in politics only requires ambition (campaigning, self-promotion, persistence), but success in government needs applied experience, good judgement and intelligence. Worth remembering when we consider our UK political class.
5
0
69
āDo not dwell in the past, do not dream of the future, concentrate the mind on the present moment.ā - Buddha Focus on whatās infront of you.
6
1
74
We evolved sentience as an adaptive feature - pain, pleasure and then the associated biological hardware; with extremely integrated recursive/recurrent processes. Not a stretch to think AI will not achieve this, and therefore, we must be extremely cautious in how we develop
What if AI is already conscious? Sentience explained | LSE Research https://t.co/01xxQ2SaSr "The troubled history of misjudging sentience in humans and animals We have a very bad track record in dealing with cases of uncertain sentience. That, for example, there was often
25
8
221
If youāre hunting for good news hereās something: Talkspace therapy is covered by most insurance plans and the average copay is only $15 (but most covered members pay $0).
33
103
855
Thereās a rising affinity for authoritarianism thatās been driven by disdain for failed liberal policies, polarisation in opinion, and an increasing incapacity for civil discourse.
7
0
89
Charlie Kirk has been shot. Remind me who the violent ones are again. Politics is broken.
8
0
76
When someone shows you who they actually are, believe them. Nobody changes.
4
0
77
Turn data into actionable insights with IDI Consultingās Analytics services. We help you unlock the power of your data to drive smarter decisions and achieve better outcomes. Contact us today to get started. https://t.co/FnmE78EEuP
6
15
215
The biotech-AI revolution can cure disease and extend life, but only if we can share health data more effectively while protecting truth and privacy.
7
2
92
The more someone insists theyāre honest, the less likely they actually are. Same goes for virtue signalling - almost every time.
9
0
74
Thereās survivorship bias in investing. We hear from the big winners, not the silent majority who lose or suffer from a high risk tolerance.
9
1
80
We are live on the @MarioNawfal Roundtable! Join @deeban_r @ceobillionaire @Kellykellam @buzzdefi0x @chrisjterry @cotlage and others in navigating the rapid progress of decentralised AI, why it matters and how web3 as we knew it has changed!
6
15
94
š¬ Hoping to attract more people who are interested in doing good work with us. Primary focus right now: https://t.co/oHVbfQMNDd āŖļø ā Academic work with @KingsCollegeLon and @NurOpus (thereās a tremendous amount of research that needs to be done, esp in AI safety)
14
13
152
The majority of LLM benchmarks are fixed questions, single-pass at temp=0 --> hiding instability, while only a handful sample statistically for harm, deception, refusal.. š§ Monte Carlo!
6
2
79
AI should act in our best interests. Beyond outcomes and benchmarking, but deeply in how it "thinks". Many models already show strategic deception before telling us what we want to hear. Thatās not acceptable at population-scale deployment.
@deeban_r What do you mean by aligning "AI with human"
12
1
73
Seneca ended each day with 4 questions: What bad habit did I curb? How am I better? Were my actions just? How can I improve? Timeless self-audit.
3
1
132
On human learning, creativity and intelligence. Real learning isnāt just about memorising a method or repeating a fact. Itās about building mental models, structured internal maps of a concept viewed from multiple angles. Thatās what allows you to move through an idea in your
12
4
118
Career politicians arenāt fit-for-purpose. Thatās the problem with the UKās political class. Detached from reality and actual business.
11
3
118