pottspotts

Yeah this is oft-repeated by particularly good "lay" engineers.

How does a man even get to this point?

I will tell you how: derangement, a failure to engage with society and the degeneration/depravity of being one step away from making MIDI component improvised explosives and emailing your remaining friends a manifesto.

Ah just kidding, this is pretty sweet.

The car industry has been lobbying congress and locales for 50+ years. Laws like jaywalking were at the behest of car companies, and that alone makes walking legally very difficult in nearly any area with a downtown.

The lack of subsidies certainly don't help. Neither does the insatiable appetite for new cars.

We think of ourselves more intelligent than the generation before us, and more wise than the ones after.

I assume this was 4o? Whenever someone says GPT would be "useless" at the given task, I think they've only tried it with older/dumber models. Almost without fail 4 seems to get the answer right.

I would argue that the neutrino is relatively large compared to the quanta typical of these kinds of tests, but in the abstract they suggested a basketball!

This is a really great experiment. It's nice to see real work being done on the ground with these kinds of tests and write-ups. I am not a physicist but I imagine this could have implications for gravitational wave detection by improving LIGO's optical sensors, maybe optical sensors in general?

Or maybe this could help with tracking time more accurately? Hopefully someone with knowledge can chime in with what this means in practice.

I came here to make this comment as well.

This line of reasoning that LLMs "only predict" the next token is akin to saying humans can only think or speak one word at a time. Yes, we use one token/word at a time, but it is the aggregate thought that matters, regardless of what underlies it.

How do you "fill something with vacuum"?

Hilariously, when I got to the end of that article, I noticed he had a theme selection dropdown hidden away in the bottom right corner.

You may suffer from some color blindness. There are easy tests online to confirm.

After some checking, it seems to be an issue with the free Heroku instance they are running.

I will miss the ease of use of Google Domains. I fear the new Squarespace domain management site will be rife with confounding settings and lit up with various upsells. They must have paid a handsome price for this.

I strongly disagree with Cal on most of his points here; he effectively recapitulates the "Its just guessing the next token" argument. In his example, he explains that ChatGPT couldn't possibly understand how to remove a PB&J from a VCR, but asking GPT what to do discredits this analogy almost entirely. It is true that GPT doesn't have servos with feedback or other hardware to remove a sandwich itself, but it can certainly explain the process including how to clean the leftover residue and follows the details by suggestion you seek out a professional technician. In the same article he admits ChatGPT is immensely complex and that "any number of humans" could not reproduce the rules by which it works; but then goes on to assure the reader that is naive and is not a threat to their jobs or safety.

This seems really disingenuous to me given he's a professor of computer science at a respected university. This kind of cynicism, that you might call naive, leads readers to put aside concerns about safety.

I strongly disagree with Cal on most of his points here; Cal wrote an article in the New Yorker yesterday that effectively recapitulates the "Its just guessing the next token" argument. In his example, he explains that ChatGPT couldn't possibly understand how to remove a PB&J from a VCR, but asking GPT what to do discredits this analogy almost entirely. It is true that GPT doesn't have servos with feedback or other hardware to remove a sandwich itself, but it can certainly explain the process including how to clean the leftover residue and follows the details by suggestion you seek out a professional technician.

In the same article he admits ChatGPT is immensely complex and that "any number of humans" could not reproduce the rules by which it works; but then goes on to assure the reader that is naive and is not a threat to their jobs or safety.

This seems really disingenuous to me given he's a professor of computer science at a respected university. This kind of cynicism, that you might call naive, leads readers to put aside concerns about safety.

For a significant number of software developers, GPT and Github's Copilot have replaced StackOverflow, and even Googling more generally. It is more than an autocomplete, it is the best resource for software development by far, IMO. It's a tutor that's an expert in virtually every topic.

The goal posts for AI are moving quickly, and in my mind, a lot of the criticism os too shallow.

People want it to perform better than any expert human at any possible subject before it's considered "real AI". It isn't enough for critics for it to be better than the average person at virtually everything its put to the test on.

It seems like there is some resentment and almost anger at this technology, particularly with the artistic AIs like Midjourney. I can understand that more readily, but what's the real beef with ChatGPT?

I am surprised by how many, even among the tech community, wholesale disregard GPT as a glorified auto-complete, or "a statistical model on human information".

What, then, is the human brain if not a trained statistical model? Granted it is considerably more sophisticated in some ways, but in many other ways it is less sophisticated and less capable.

We are moving the goal posts on AGI very quickly, but it is catching up. I think we need to appreciate the nature of this milestone if we have any hope of controlling potential singularities.

Perhaps its time we pay for a service like ChatGPT for an ad-free experience. I'm actually already on the $20/mo plan.

There's a lot of focus on the "bullshit generation", but this problem is not unique to LLMs. I find it much harder to cut through the bs of search engines. A lot of times the first page of a Google result is ads, including maps/etc.

It's gotten a lot worse than 10 years ago, and then there's the SEO spam the other comment mentioned.

I find ChatGPT to be a lot more effective for work as a developer, and I've essentially stopped using Google search day-to-day.

I've largely used GPT as a replacement for Google, and I find it to be a lot smarter and faster. The criticisms the author level at it don't seem based on having used it for anything substantial.

It is an excellent tool and a major step, and it will only get better and easier to use, it seems.

This is really great, thank you. I would love to see a real "from scratch" that doesn't use torch.py et. al., though.

Are you taking credit for this? I have seen this Matrix before, posted by the original author, and you should credit him.

In any case I agree with the majority here that this tool is akin to standardized testing which at long last is seen for the bad tool that it is for finding objective quality.

About 10 years ago my friends and I made a Web site, removethepenny.com (no longer active) and starting calling around to see who would support us. I recall speaking to someone who was very angry that we cared so little for the metal industry.

I think this has been known for a long time but special interests have kept it on life support for obvious reasons.

I like this idea. I have also been bad about picking podcasts that I like, though. I think laziness and procrastination might be a huge factor for me. :-?

It may have helped that he became more proficient at dieting and using the app as well as intermittent fasting.

This is incorrect. The laws of physics in play are biochemically complex but the net energy differential is straight forward. Calorie Intake > Burn Rate = Weight Gain, and vice-versa. Exercise increases the burn rate, but it doesn't change the formula.

Really great data! Thank you for sharing.

The hard part for me is mindset and willpower. Unfortunately, knowing the physics* behind weight loss doesn't make it any easier, and perhaps might make it more stressful. Why can't I do this simple thing that logically is as complex as 2+2. Our minds have their own prerogatives.

I wonder then what did you tell yourself? How many times did you have a bad day? What was your mood like throughout? I'd love to know so much more about the psychology. Great article though and thanks again for sharing!

* https://www.youtube.com/watch?v=vuIlsN32WaE

This is very interesting. It is important to remember that "man-made" and "natural" are on the same continuum. We are a part of nature, and vice-versa. This discovery is stunning, but not surprising, given the other incredible inventions by mother nature in the history of evolution.