padolsey

9 ms·

91.
by padolsey 1y ago
This gives me so much joy. Well done!!
92.
Sorry, We Deprecated Your Friend(blog.j11y.io)
4 points by padolsey 1y ago|2 comments
93.
by padolsey 1y ago
Ugh Anthropic please don't become the villain.
94.
by padolsey 1y ago
This is a huge shame. Browsers are one of the more ideal sandboxing barriers. The likes of Chrome and Firefox could have worked with OSs and AI labs to ensure a more robust system of mitigations were in place. Setting legitimizing precedent
95.
by padolsey 1y ago
This is why I've found the safety research conducted by the likes of Anthropic and OAI to be so confusing. Like when they said that models are likely to blackmail developers in order to avoid being 'turned off' [1]. What an u
96.
by padolsey 1y ago
Hi jph! Not intending to subvert the thread, but I'd love to chat to someone like you. The non-profit I work at has been working on democratizing evals. This wouldn't be to ensure your in-house AI is up to scratch (parachute look
97.
by padolsey 1y ago
You're right; we should not even try. Better to have 0% compliance coverage and your honor than 90% coverage as best-effort.
98.
by padolsey 1y ago
> This is cool, but I’m a little skeptical. If Parachute uses AI agents to evaluate other models, who’s evaluating the AI agents? Usually you can run human-in-the-loop spot checks to ensure that there's parity between your LLM evalu
99.
by padolsey 1y ago
Most of these attacks succeed because app developers either don’t trust role boundaries or don’t understand them. They assume the model can’t reliably separate trusted instructions (system/developer rules) from untrusted ones (user or
100.
by padolsey 1y ago
Reminds me of haikus; to be true in nature, they must have a 'cutting word' to severely juxtapose, allowing two otherwise irreconcilable meanings to be bridged. A good haiku must be composed from two directions, not one. But LLMs
101.
by padolsey 1y ago
> But I will ask hard questions to see how well a candidate communicates in a stressful situation Well, stress fires off old cortex fight-or-flight response. It's like the worst possible test of neocortex capability. Why are you try
102.
by padolsey 1y ago
> Some companies genuinely care about this. Some even mention it in their job descriptions. They want candidates who perform well under pressure. Who are these self-important employers?? I mean, outside of Ops and live incident response,
103.
by padolsey 1y ago
>> LLM's do not possess professional experience needed for successful therapy, such as knowing when to not say something as LLM's are not people. > Most people do not either. That an LLM is not a person doesn't seem
104.
by padolsey 1y ago
Entirely anecdotally ofc, I find that therapists often over-bias to formal diagnoses. This makes sense, but can mean the patient forms a kind of self-obsessive over-diagnostic meta mindset where everything is a function of trauma and fundam
105.
by padolsey 1y ago
Ofc the title got its point across, but I'd argue to hold ourselves to higher standards of veracity. That's all.
106.
by padolsey 1y ago
Yep it's always a bit cute and funny, when you consider the absolute necessity of dopamine in basically every functionally relevant neural activation. Talk to a parkinson's patient about your 'dopamine=bad' fluff. Ugh. T
107.
by padolsey 1y ago
> Using ChatGPT on cognitive tasks can reduce your brain connectivity by up to 50% and reduce your ability to recall information about the task by 8x. Argh people keep referencing this study as Gospel. It has not been peer-reviewed. Its
108.
by padolsey 1y ago
Agree utterly. It's a real shame, and severely affects accessibility for disabled and elderly people. Not only UI discoverability but also the types of swiping or holding movements required on mobile devices. The initial mobile interfa
109.
by padolsey 1y ago
A know a person at the FCDO who had to routinely write letters in correspondance to those who'd sought Palmerston's advice on various matters. A hilarious internship.
110.
by padolsey 1y ago
I really recommend people study the measurement frailties and prompting sensitivities of LLM judges before employing them. They're valuable, but should be used with complete understanding of the risks: https://www.cip.org&#x
111.
by padolsey 1y ago
Agreed! FWIW I am attempting to create an open-source wiki/watchdog eval platform -- weval.org -- , so we can all keep an eye on LLMs, their biases, and their general competencies without relyong in the AI providers marking their own h
112.
by padolsey 1y ago
How much transparency does Claude Code give you into what it's doing? I like IDE-integrated agents as they show diffs and allow focused prompting for specific areas of concern. And I get to control what's in context at any given t
113.
by padolsey 1y ago
IMO This is such disingenuous and misleading thing for Anthropic to have released as part of a system card. It'll be misunderstood. At best it's of cursory intrigue, but it should not be read into. It is plainly obvious that an LL
114.
by padolsey 1y ago
If we're paying for reasoning tokens, we should be able to have access to these, no? Seems reasonable enough to allow access, and then we can perhaps use our own streaming summarization models instead of relying on these very generic-s
115.
by padolsey 1y ago
> if you want to sculpt the kind of software that gets embedded in pacemakers and missile guidance systems and M1 tanks—you better throw that bot out the airlock and learn. But the bulk of us aren't doing that... We're making C
116.
by padolsey 1y ago
I feel the same. On reflection, it's how I think I experience thoughts emerging in my head. Language gets derived from initially noisy embeddings. It's quite beautiful that we've ended up closer to a de-noising architecture t
117.
The people stuck using ancient Windows computers(bbc.com)
9 points by padolsey 1y ago|3 comments
118.
by padolsey 1y ago
Interesting! > People who don't pause exist more in their head than their body. The mind is top-down, rigid, quick, enforcing an established view. The mind is waiting for the other person to be done so they can say what’s rattling a
119.
The CEO of Fiverr says "AI is coming for you"(entrepreneur.com)
5 points by padolsey 1y ago|2 comments
120.
by padolsey 1y ago
> That's just an advertisement for their competitor. I think the simpler explanation is just laziness and no positive incentive or obligation, as opposed to proactive competitive practices..

More ›