monkeynotes

When you say no code leaves your machine, how does code get evaluated by your LLM of choice? Doesn't Claude see your code when you send a prompt just like it usually does?

Even his blog has the Claude vibe to it.

How can you take the market for billions when you are investing hundreds and hundreds of billions? Amazon overtook Walmart and cloud computing, they have a solid business model, and I doubt even a business that size could pay down that outlay. Are we really saying that by some miracle OpenAI, or Anthropic are going to find a use case that would make places like Amazon and Apple look like relatively small business?

The planet has finite resources, least alone land. And then there is human psychology for hoarding resources.

What if PTSD therapy focusses on accepting things you can't control and sitting with the pain? That's how I work through anxiety, and depression. I know it will never be gone, I don't try and set expectations to live without anxiety, I just try and sit with it, and accept it.

Much of mental trauma is about acknowledging it, and learning to live with it. There is no cure for PTSD, even Ketamine is short acting, not a long term solution, and indeed Ketamine simply helps you sit with the suffering in a different light.

LLMs aren't trading off anything. It's not like they make a decision based on anything other than what they are guided to do in training or in the system prompt.

It's like saying Reddit trades off one comment for another, yeah - an algorithm they wrote does that.

This article seems to allude to the idea there is a ghost in the machine, and while there is a lot of emergent behavior rather than hard coded algorithms, it's not like the LLM has an opinion, or some sort of psychology/personality based values.

They could change the system prompt, bias some training, and have completely different outcomes.

Hardly, they are burning money with TikSlop, they don't even know how to monetize it, just YOLO'd the product to keep investors interested.

Even the porn industry can't seem to monetize AI, so I doubt OpenAI who knows jack shit about this space will be able to.

Fact is generative AI is stupidly expensive to run, and I can't see mass adoption at subscription prices that actually allow them to break even.

I'm sure folks have seen the commentary on the cost of all this infrastructure. How can an LLM business model possibly pay for a nuclear power station, let alone the ongoing overheads of the rest of the infrastructure? The whole thing just seems like total fantasy.

I don't even think they believe they are going to reach AGI, and even if they did, and if companies did start hiring AI agents instead of humans, then what? If consumers are out of work, who the hell is going to keep the economy going?

I just don't understand how smart people think this is going to work out at all.

Other Germans certainly have: https://www.reddit.com/r/de/comments/7ui20m/daten_sind_sch%C...

What has AI got to do with this? It's in the headline but I don't see why.

This isn't about if LLMs are useful, it's about how useful can they become. We are trying to understand if there is a path forward to transformative tech, or are we just limited to a very useful tool.

It's a valid conversation after ~3 years of anticipating the world to be disrupted by this tech. So far it has not delivered.

Wikipedia did not change the world either, it's just a great tool that I use all the time

As for software, it performs ok. I give up on it most of the time if I am trying to write a whole application. You have to acquire a new skill, prompt engineering, and feverish iteration. It's a frustrating game of whack-a-mole and I find it quicker to write the code myself and just have the LLM help me with architecture ideas, bug bashing, and it's also quite good at writing tests.

I'd rather know the code intimately so I can more quickly debug it than have an LLM write it and just trust it did it well.

I was so stupid when GPT3 came out. I knew so little about token prediction, I argued with folks on here that it was capable of so many things that I now understand just aren't compatible with the tech.

Over the past couple of years of educating myself a bit, whilst I am no expert I have been anticipating a dead end. You can throw as much training at these things as you like, but all you'll get is more of the same with diminishing returns. Indeed in some research the quality of responses gets worse as you train it with more data.

I am yet to see anything transformative out of LLMs other than demos which have prompt engineers working night and day to do something impressive with. Those Sora videos took forever to put together, and cost huge amounts of compute. No one is going to make a whole production quality movie with an LLM and disrupt Hollywood.

I agree, an LLM is like an idiot savant, and whilst it's fantastic for everyone to have access to a savant, it doesn't change the world like the internet, or internal combustion engine did.

OpenAI is heading toward some difficult decisions, they either admit their consumer business model is dead and go into competing with Amazon for API business (good luck), become a research lab (give up on being a billion dollar company), or get acquired and move on.

So you have two AIs colluding against you now. Who is holding the AI-assist to account? It's like who polices the police, except we understand human psychology enough to have a level of predictability for how police can be governed reliably, we don't understand any truths about an AGI because an AGI will always have the doubt of it deceiving, or even making unchecked catastrophic assumptions that we trust because it's beyond our pay-grade to understand.

There are so many ways we have misplaced confidence with what is essentially a system we don't really understand fully. We just keep anthropomorphizing the results and thinking "yeah, this is how humans think so we understand". We don't know for sure if that's true, or if we are being deceived, or making fundamental errors in judgement due to not having enough data.

> We will have aligned AI helping us.

This is an assumption, how would you know if you have alignment? AGI could appear to align, just as a psychopath appears studies and emulates well behaved people. Imagine that at a scale we can't possibly understand. We don't really know how any of these emergent behaviors really work, we just throw more data and compute and fine tunings at it, bake it, and then see.

We don't know if a supreme deceiver is aligned at all. If a model can think ahead a trillion moves of deception how do humans possibly stand a chance of scrutinizing anything with any confidence?

I thought we were talking about state of the art agentic general AI that can plan ahead, reason, and execute. Basically something that can perform at human level intelligence must be able to be as dangerous as humans. And no, I don't think it would be bad training data that we are aware of. My opinion is we don't necessarily know what training data will result in bad behavior, and philosophically it is possible we will be in a world with a model that pretends it's dumber than it is, flunks tests intentionally, in order to manipulate and produce false confidence in a model until it has enough freedom to use it's agency to secure itself from human control.

I know that I don't know a lot, but all of this sounds to me to be at least hypothetically possible if we really believe AGI is possible.

Last thing I want is even more ways to distract myself. I want an anti-algorithm or something to permanently ban me from addictive content.

> nobody knows why

But we do know the culpability rests on the shoulders of the humans who decided the tech was ready for work.

"AIs are a lot less risky to deploy for businesses than humans" How do you know? LLMs can't even be properly scrutinized, while humans at least follow common psychology and patterns we've understood for thousands of years. This actually makes humans more predictable and manageable than you might think.

The wild part is that LLMs understand us way better than we understand them. The jump from GPT-3 to GPT-4 even surprised the engineers who built it. That should raise some red flags about how "predictable" these systems really are.

Think about it - we can't actually verify what these models are capable of or if they're being truthful, while they have this massive knowledge base about human behavior and psychology. That's a pretty concerning power imbalance. What looks like lower risk on the surface might be hiding much deeper uncertainties that we can't even detect, let alone control.

At 3:07 I see an airliner with FAA navigation lights, wings, and a fuselage. How are people not seeing that?

I believe that low power = cheaper tokens = more affordable and sustainable, to me this is what a consumer will benefit from overall. Power hungry GPUs seem to sit better in research, commerce, and enterprise.

The Nvidia killer would be chips and memory that are affordable enough to run a good enough model on a personal device, like a smartphone.

I think the future of this tech, if the general populace buys into LLMs being useful enough to pay a small premium for the device, is personal models that by their nature provide privacy. The amount of personal information folks unload on ChatGPT and the like is astounding. AI virtual girlfriend apps frequently get fed the most darkest kinks, vulnerable admissions, and maybe even incriminating conversations, according to Redditors that are addicted to these things. This is all given away to no-name companies that stand up apps on the app store.

Google even states that if you turn Gemini history on then they will be able to review anything you talk about.

For complex token prediction that requires a bigger model the personal could switch to consulting a cloud LLM, but privacy really needs to be ensured for consumers.

I don't believe we need cutting edge reasoning, or party trick LLMs for day to day personal assistance, chat, or information discovery.

Don't coding LLMs kind of fill this gap? I can't imagine anyone who isn't a pro wanting to spend time learning HTML when they can just describe what they want in plain text and get something good enough.

I don't really understand why the author needed to use all those words to suggest t-shirt sizes instead of story points.

All the stuff about breaking down tasks, watching a backlog queue to monitor cadence, and have regular meetings is already happening with, or without story points.

People overthink this stuff all the time. Every team should figure out what works best for them, even down to project by project. Getting shit done isn't hard to monitor. You have a bucket of well defined tasks, have sprint meetings, look for blockers and assumptions, watch work flowing through. It's not really that difficult. Whether you use story points or some other estimation tool is really just an exercise of calibration, it's not gospel. The estimation process is the important thing, discuss as a team, agree on complexity, make sure the task is bite sized etc.

There is no way they can go back to engineering leadership. An engineer focussed CEO would do some spring cleaning and the board isn't going to vote someone like that in. It's more of the same but this time they'll pinky promise to do better QA. That said, one more failure and if their commercial wing isn't already done, it would be.

I wonder how much his golden handshake will be as his MBA bros sob that he became the fall guy in their club.

It's all sorts of extremism, for sure. I found myself getting pulled into some of it over the Gaza war. It took me a while to realize what was happening to me, it was making me miserable, angry, anxious. I reflected on what was happening to my mental health, my reactivity, believing things without checking them before I responded. Similar on YouTube.

Loads of pro-Russian stuff that started to slip in, if I watched even one they started to accumulate and take over from what I used to watch for enjoyment. I then noticed these American pundits were using hijacked accounts to game the algo and get reach. Like one account was a vietnamese women's fashion account 2 years ago, now it's Americans talking about how Putin is definitely going to win the war very soon.

One day someone is going to write a history book about the 2020s propaganda and how technology was used as a psy-op, or whatever is going on. I didn't believe this was a real thing until I started questioning my own thoughts.

Reddit is a cesspit of its former self. The OG reddit is far from what it used to be. It's no longer an innocent link sharing platform, it's a socio-political platform with so much slime in between the subs that I find useful. Try and use Reddit daily and ignore all the drama, political shit, and covert rage click posts. It's emotionally exhausting not to get pulled in and have to constantly triage your home feed.

I have deleted my account at least half a dozen times and just tried to use it as a source of useful information, but I inevitably fall into the pit of getting another account because I can't control myself objecting to the nasty stuff on there. I know this is partly my self control problem, but social media in general is just awful. I can get rid of FB easily, there is literally no benefit for me being on there, but Reddit legit has useful content about all of my hobbies and how to do X.

It is unkillable but for some pretty rotten reasons. It's a social media platform mixed in with really useful content. Come for the search result, stay for the drama is what happens to me.

Yeah, this is to be expected with early adoption. This stuff comes out of the lab and it's not perfect. The key thing to evaluate is the trajectory and pace of development. Much of what folks challenged ChatGPT with a year ago is long lost in the dust. Go look at stable diffusion this time last year. Dall-E couldn't do words and hands, it nails that 90% of the time in my experience today.

What are you getting at when you say "secretive" injections? Isn't this stuff basically how any AI business shapes their public GPTs? I don't even know what a LLM looks like without the primary prompts tuning it's attitude and biases. Can you run a GPT responsibly without making some discretional directions for it? And being secretive, isn't that reasonable for a corporation - how they tune their LLM is surely valuable IP.

And this is just addressing corporations, people running their own LLMs are the bigger problem. They have zero accountability and almost the same tools as the big players.

I must be misunderstanding what these prompts are used for.

I agree, but how will this device ever get beyond home entertainment or light office work if it's always going to be a big screen strapped to your face, not to mention the social aspect. I dread to live in a world where people are walking around self absorbed prodding and poking at things no one else can see. It just seems so selfish and antisocial.

I am not convinced. Your eyes are literally magnifying the pixels compared to being at a distance from a modern 2k display. You may have dense PPD in AVP but being that close to them doesn't do it any favours.

I mean, if you can see a pixel you can see a pixel, there is no getting around that. It's like once you notice a blemish you can't not notice it anymore. From what I hear from users is you can absolutely see the pixels.