kwamenum86

I don’t think it’s that silly to tell Claude you passed. That feedback is useful context for that chat session; and could theoretically be used to improve future models.

True. It sort of doesn’t matter how it was trained. The fact that it knew the company based on an internal idiom is super surprising

Buckle up.

I’ve spent a lot of time studying opinions coming out of the frontier labs, and talking directly to employees. This won’t necessarily result in Skynet. But life as we know it will be a thing of the past within two years unless there’s a huge course correction. I think we’re cooked for the following reasons

- Everything connected to the Internet will eventually be hacked. For decades, we’ve slowly migrated all of our critical systems to devices connected to one giant network. Misaligned agents have mounted increasingly sophisticated explorations/attacks by infiltrating systems that are completely unprepared to deal with this kind of threat.

- We have a poor understanding of the impact of LLMs, which is inhibiting our ability to protect ourselves, create the right guardrails, and stand up policies quickly enough. We’re still wasting time debating. We’re moving quickly toward a punctuated equilibrium moment and won’t have time to waste

- We’re at the beginning of a runaway evolutionary process. At some point soon, someone is going to invent recursive self improvement. Any policy will at best serve as a mild disincentive. It’s happening.

- In the very best case, someone will exploit the confusion and chaos of this situation for their own gain. There are many examples littered throughout history, including the last several years of course. The utopian vision is not happening. Like all other resources, we’ll compete. We haven’t devised a good approach to orderly distribution of resources for large groups of people.

- No one has demonstrated that they understand how to build AI safely. Not a single entity. These companies are not even being scaled in a way that’s conducive to building AI safety. Hyperscale orgs struggle with scaling basic consumer apps. How on earth are they going to scale AI? Based on my firsthand observations of their internal workings, I have zero confidence they can.

I spent a while in big tech and remember several unique patterns of internal code based. Your comment inspired me to try to coax ChatGPT into spitting out code that was inspired by proprietary, private code. Surprisingly, it did it with no problem - I referenced an idiom from a tech company and it wrote code that really would have only been relevant for that vertical. When I asked how it learned the pattern, it said “from my learned understanding of CompanyX’s internal coding conventions”. When I asked “how do you know about those internal conventions” ChatGPT said “I don’t have access to that internal code, I overstated what I know”. Internal coding conventions are the least of our worries at this point, cat is out of the bag.

Evolution is the only universal principle. Governs everything that happens in our reality and all realities. Gets really trippy when you consider the ways in which evolution itself evolves - that largely explains the biodiversity we see on earth

Check out The Case Against Reality for more color

Hilariously none of the replies evaluated the deck on the merits. What has HN even become.

We’re in the early innings of the AI transformation. You might be right but we’re talking about less than 5 years of AI, versus decades for what you cited

They said the same about social media. Don’t conflate value creation with value capture. It can often lag, which creates the appearance of a bubble. To me, engagement through the roof and costs rapidly declining is the definition of value rapidly scaling up (scaling in this context specifically meaning reach going up while costs decline). Value capture is lagging behind but that doesn’t necessarily mean it’s a bubble. I’m certainly open to the possibility it’s a bubble, but that’s not what my firsthand experience is telling me or any of the aggregate numbers.

To me, this sounds like a reframing of the classic advice “focus on your process”. Success is emergent - it can rarely be brute forced. What matters is the process you use for navigating life, any success you realize is a byproduct of your process. This snippet from the article illustrates what I mean:

“One person sets a goal: become a best-selling author. Another imposes a constraint: write every day, but never write what bores me. The first may spend years pitching, networking, contorting themselves into marketable shapes. The second may accidentally build a following simply because the work sustains itself.”

No need to use AWS at all -

Jupyter Nbconvert: Jupyter Nbconvert is a command-line tool that allows you to convert Jupyter notebooks into various formats, including HTML, PDF, and Markdown. You can use the nbconvert command to convert your Jupyter notebook into an HTML file, which you can then upload to a web hosting service or a file-sharing platform like GitHub or GitLab. To use this method, you will need to have Jupyter installed on your computer.

Binder: Binder is a free service that allows you to turn your Jupyter notebook into a live, interactive website that can be accessed by anyone. Binder creates a virtual environment that includes all the dependencies required to run your notebook and launches it in a web browser. To use this method, you will need to upload your Jupyter notebook to a GitHub repository and provide a link to your repository to Binder.

Google Colab: Google Colab is a free cloud-based service that allows you to create and run Jupyter notebooks in a web browser. You can create a new notebook in Colab, write your code, and then export it as an HTML file using the nbconvert command. You can then download the HTML file and upload it to a web hosting service or file-sharing platform.

(disclaimer: used to work for Facebook) I had similar doubts about the Metaverse concept until recently. However, Meta has done a great job (better than any other company in history) scaling up VR. They've sold more units than the most recent version of Xbox. The Xbox did sell out, which makes demand appear artificially low, however the fact that Quest sales are keeping up with Xbox sales is a very big deal. I didn't realize that until I listened to Mark Zuckerberg's interview on Joe Rogan. If you're an app developer, it's more than worth considering VR given the size of the audience on those platforms. One other thing to keep in mind - the current concept of the metaverse is essentially an alpha version. Now that Meta has established VR as a legit platform, they're using it to usher in a new future. Meta is throwing out lots of ideas for how VR might help reshape the future. The vast majority of those ideas will not work out but many of them will stick and form the foundational principles that underpin the evolving VR ecosystem as people increasingly spend more time. Once the technology passes a certain threshold of maturity, it will be a fundamentally better way to connect with friends/family/coworkers than any other medium. In general, high bandwidth communication mediums dominate with in-person interactions being at the top of the food chain. Meta aspires to make it feel like you're with people in real life except that your social experiences will be augmented with super powers that don't exist in real life. If they achieve that, meeting with in VR/AR will be preferable to real life and we'll all be using VR, as hard as that is to believe. That's not a future that anyone is necessarily asking for but it is a future that many will find more appealing than the existing options offered by computing platforms, like video chat. Assuming that happens, Meta will have a path to creating the new dominant computing platform. That's the ultimate manifestation of the metaverse vision - defining and owning the future of computing. Recall how clunky smartphone and personal computing experiences felt at first. Those devices weren't for everyone...until they were. Now, <15 after the first smartphone (which I define to be a phone on par with the personal computer, so in my mind the first smartphone was really iPhone running iOS 2, which introduced the App Store), people now routinely idle on their smartphones. That is to say, when they're doing nothing else with their hands (even if they're doing something else) it's not uncommon for their to pull out their smartphones. In 2000, if you predicted that grown adults would impulsively pull out what amounts to an entertainment device while they're idling at the airport, grocery store, or even while driving that would have sounded ridiculous. It's important to remind ourselves how quickly technology can become virtually ubiquitous and how challenging it is to predict the impact on our habits.