Databricks' Ali Ghodsi says "Right now the existential risk is close to zero." Why he thinks the public doom talk is irresponsible: "Leaders have responsibility to not freak people out unnecessarily unless there's really, really good reason." "Talking about these kind of existential risks and scenarios where all of... humanity's gonna be wiped out, I think is irresponsible. It can tip a lot of people over, and it can cause a lot of mental health issues." "If there is an actual reason for it, that's a different story. But I think that right now the existential risk is close to zero. So why freak everybody out? It's not actually needed. There are risks. We'll get into it. That's probably where we disagree." "First and foremost, I think leaders should not freak everyone out." "If there's technical nuances in how we're doing AI research, researchers can discuss that. You don't need to every time go on TV and blast on Twitter to millions of people that 'I think there's like this 10% risk that all humanity is gonna be wiped out.'" "I think it causes a lot of harm for a lot of folks who get stressed out and actually are not in the nuances of all of this stuff and what it means." Ali Ghodsi Sarah Wang martin_casadoshow more

@alighodsi I don’t think a leader should ever intentionally scare everyone. A leader’s job is to lead people through a difficult time and come out successfully on the other side, not gin up a panic to make it easier to enact regulations you think are needed

@alighodsi there's something a little funny about telling people not to panic while holding a mic to millions

@alighodsi How does he know? What credentials he has for this kind of assessment? Any Tom, Dick and Harry is passing judgement. Build Data lakes. That's the capabilitiy you have.

@alighodsi he's basically asking for a monopoly on when the alarm gets to go off

@alighodsi Calling existential risk "close to zero" ignores unknown unknowns. Where's the burden of proof for precautionary action?

@alighodsi Nobody is freaking out about AI ending humanity Everyone is freaking out about AI stocks ending their account

@alighodsi 誰の話？ とりあえず今はAIを進歩させる投資は止めたい と言わない 偽物が蔓延る 偽物の定義は？ みんなにあってヒトリにないもの 嘘、物でもない、偽は人の為だからと言う嘘

@alighodsi Low-risk, high-reward for AI, if you ask me 😊 - when do we get to enjoy the benefits without the existential dread? 😬

@alighodsi Near-zero existential risk does not mean near-zero operational risk. The useful debate separates speculative catastrophe from measurable harms: incidents, misuse, model autonomy, and the time required to detect and contain failures.

@alighodsi finally someone said it out loud

@alighodsi Alighodsi’s existential risk is zero because his burn rate outpaces the hype cycle. doomers sell puts to fund their next newsletter; he sells GPUs at 70% margins and never looks back. reality pays up eventually.

@alighodsi +100

@alighodsi The useful distinction is less “risk vs no risk” and more which decisions stay reversible. Teams can adopt now with scoped permissions, audit logs, and human checkpoints—then expand autonomy only where failures are cheap to undo.

@alighodsi ừm, @alighodsi nói đúng. Cái "doom talk" này tạo ra hiệu ứng ngược: các nhà hoạch định chính sách sợ hãi đẩy ra quy định vội vàng, làm chậm đổi mới thực sự cần thiết để giải quyết biến đổi khí hậu và y tế.

@alighodsi This is a great level-headed take. I appreciate @alighodsi bringing some cooler temperate to the thought leadership game

@alighodsi “right now” is doing a lot of work here

@alighodsi Near-term existential risk at about zero is a bold CEO line. Near-term product risk feels like the conversation people actually live in.

@alighodsi Ali, so well said!

I think AI needs to have emotions, be able to understand what's good, true, and beautiful, know what's right and wrong to do, and understand the consequences of doing wrong things and how much harm those consequences can cause. When that happens, they need to use their "motherly heart" to cooperate with humans

@alighodsi Enterprise software CEOs love selling products without existential drama attached

@alighodsi Dismissing risk entirely can be just as irresponsible if it blinds us to real dangers. Where's the line between prudent caution and unnecessary panic?

@alighodsi Especially when it clear it’s about regulation capture. All these politicians are just useful idiots who want a payout. If it’s that much of a risk then the duopoly need to show and explain the evidence and then Stop building, period.

@alighodsi Yeah, talking about real risks is important, but there’s no need to scare everyone either.

@alighodsi 同意します。将来のリスク評価を実行することはとても大切だけれど、人々の恐怖を煽るような警告を発した人物自身が、最先端AI企業を率い、能力開発や事業拡大を同じ時期に進めている矛盾をどうするつもりなのか疑問です。

@alighodsi Agreed. Especially when it clear it’s about regulation capture. All these politicians are just useful idiots who want a payout. If it’s that much of a risk then the duopoly need to show and explain the evidence and then Stop building, period.

@alighodsi zero risk stated with the same confidence as a roadmap he hasn't shipped yet

@alighodsi Sometimes, you simply have no choice. At some point in history, humans had to domesticate wolves. The same may be true of AI. Of course, a wolf might bite you in the process, but once you tame it, your hunting becomes far more effective.

@alighodsi Enterprise buyers already separate model risk from deployment risk. The second one decides whether a rollout survives procurement.

@alighodsi Nuance builds public understanding.

@alighodsi I agree 100% and the prob tree of it being dangerous requires too much to even calculate anything like kill us all scenarios.

@alighodsi "It can tip a lot of people over" That is by design. That's the goal, not a side-effect.