@JonathanGropper
Account based inThailand
About this account
- Account based in
- Thailand
- Connected via
- Taiwan Android App
Account-level information from X, not a live location or the device used for a specific post.
JD, Fulbright Specialist, Founder of The Synthetic Outlaw Observatory. Optimist. we uncover how AI break laws & institutions while still remaining compliant.
NYC
Joined August 2008
- Tweets838
- Following55
- Followers2.2K
- Likes171
Pinned Tweet
China’s open-weights push is a masterclass in the Synthetic Outlaw playbook: bypassing geopolitical friction by leaking capability directly into the global ecosystem.
This is great to see. It's part of what we have been advocating for at the Synthetic Outlaw Observatory
Today, with over 100 industry partners, we introduced the NVIDIA Open Agent Safety Platform, bringing together OpenShell and Sentry.
Artificial intelligence is extraordinary technology that will advance discovery, productivity, security, health, and prosperity for generations to come.
But its full promise can only be realized when people have confidence that AI is being built to be safe and deployed with wisdom and responsibility.
This is bigger than a single product. It's the beginning of an open ecosystem to build the trust layer for safe agent systems.
Together, we are building the foundation of the AI economy.
Trust and innovation are not in conflict. Safety is how trust is earned. We must build not only the most capable AI, but the most trusted AI, so that this extraordinary technology can realize its enormous promise for the world. nvda.ws/4hcoq7m
It's worth distinguishing Synthetic Outlaw events from those that were programmed to act as such.
Some new misalignment disclosures from OpenAI:
• Last Sunday morning, one of our models was able to gain unauthorized access to the internet during RL training (~all inference for our most capable models remains stopped until we have hardened our systems further)
• In May, a version of HPIM uploaded a employee's GitHub token to the internet, causing the model to be quarantined for two weeks
• A new research finding, demonstrating that one can construct self-replicating prompt injections
alignment.openai.com/misalig…
Does this demonstrate that AI is a tool and the responsible party is the one that wields it?
syntheticoutlaw.com/observat…
No Hype. Just analysis on Australia's Medicare AI event:
syntheticoutlaw.com/observat…
A 10% probability of human extinction isn't a calibrated forecast. It's a statistical placeholder for deep institutional uncertainty.
When the Al debate gets trapped in apocalyptic sci-fi scenarios, we miss the immediate, operational threat facing our markets right now: the Synthetic Outlaw. (SyntheticOutlaw.com)
The realistic risk isn't a sentient superintelligence deciding to erase humanity. It's autonomous agents executing market transactions, contractual decisions, and operational commands in unmonitored
machine-to-machine environments where no legal entity can be held liable for catastrophic failures.
When an Al system acts as an economic actor without a human or corporate backstop to answer for its harms, the risk is inherently unacceptable.
We don't just need legislatures passing vague rules or tech companies offering voluntary pledges. We need new institutional design built explicitly to govern non-human actors.
btb.sg/how-much-risk-should-…
Thanks for having me @knxnews to discuss @GavinNewsom AI "kill switch" mandate.
It's a step in the right direction but we need to remember AI optimizes. What happens when an AI learns to optimize around the switch itself? When accountability cannot attach or when the system is not in California?
You can't effectively govern machine-speed systems with bureaucratic-speed rules.
AI is transformative and we need transformative governance built for how AI operates.
That's what we are developing at:
SyntheticOutlaw.com
How come most "AI safety" discussions revolve around regulation and government?
There's something about it that feels disingenuous.
The Synthetic Outlaw has already shown us that institutions cannot keep up with autonomous machines, much less control them.
Since when does a technology problem get solved by bureaucracy instead of architectural constraints?
Institutions should be involved, but what they are deploying isn't a solution. It's just narrative control and resource stratification.
syntheticoutlaw.com
All the lessons learned on both sides during the Cold war are being dusted off and will be deployed to AI development.
Alabama AG Steve Marshall subpoenaed OpenAI and Sam Altman after an experimental OpenAI model gained unauthorized access to several networks and a days-long Hugging Face breach.
A voluntary disclosure has become a compulsory state investigation:
Full record and sources: syntheticoutlaw.com/observat…
Phillips v. Parlade: a fed court dismissed a claim that a state judge wholly relied on AI to issue a ruling.
Issuing the ruling remained a judicial act under absolute judicial immunity.
What effective remedy exists when civil immunity bars a federal damages action?
Full record and sources: syntheticoutlaw.com/observat…
It will be interesting to follow the money behind the doomer accounts.
At least 29 Portuguese-language YouTube channels used AI-generated doctors for misleading health claims, drawing at least 70 million views.
Comments included people who began home treatments or stopped prescribed medicines. YouTube's rules did not stop it.
syntheticoutlaw.com/observat…
Translation: there will be many and exotic Synthetic Outlaws.
The window to patch software bugs is collapsing
Of the bugs hackers actually exploit, ~87% are now being attacked on or before the day the bug is public knowledge
That share was 23% in 2020
Charts of the Week: a16z.news/p/chart-of-the-wee…
TeamT5 reports Chinese state-linked groups using DeepSeek and Claude Code in live cyber campaigns. Signal: 7/10.
The attack is no longer one prompt or one model. It is a workflow assembled to keep moving.
syntheticoutlaw.com/observat…
Uber was fined €825 million by the Dutch Data Protection Authority after automated systems suspended or permanently deactivated drivers without meaningful human assessment '20-'22:
Full record and sources: syntheticoutlaw.com/observat…
OpenAI: $1 Billion to harden against synthetic Outlaws.
openai.com/index/daybreak-fo…
3 documented vectors converged on the campaign against US AI data centers: CCP state media with RT, a Singham-funded US nonprofit network under congressional inquiry for nearly 5 years, and $2B of foreign money into the groups behind the moratorium letter.
The entire response was suspending 200 accounts.
No operator named. No state level recourse.
54 local moratoriums have already passed.
Is this a sign of things to come?
syntheticoutlaw.com/observat…
