Experts Analyzed Where AI Is Heading. Here's What They Found.

[@Nanalyze]

Want to become a better investor? Signing up for our newsletter certainly couldn’t hurt 👇 www.nanalyze.com/nanalyze-weekly/?

[@businesswithredbeard5567]

We need Wyatt to make a compliation of all your backgrounds to show all the places investors can go if they invest wisely.

[@Nanalyze]

We've discussed internally (Wyatt's idea) about possibly mentioning where the videos are being shot. Since we started doing live video everything has been shot across the Pacific Northwest (Oregon/Washington/Idaho). This fall will likely be shooting across Africa so we might start calling out some of the locations briefly since people seem interested. :) Joe P.

[@WyattCounterman]

I accept this task

[@Leo-pd8ww]

Waiting for a Dutch polder background 😄

[@SirDangalot]

Hey guys, great channel and great episode. When I first saw the word "Nanalyze", I thought you were making a judgment on how truthful my grandmother may or may not be. I'm glad I clicked on the thumbnail, however, because that's not even remotely what you do. Keep up the amazing work!

[@Nanalyze]

Hah! Thank you for the comment and we'll keep it up!

[@MrMentalpuppy]

Arigatou to the amazing team for the vids! 🎉

[@Nanalyze]

Thank you for the support @MrMentalpuppy!

[@TeslaEVolution]

Even better: Domo Aregato Gosaimas : )

[@midwestcannabis]

Wow! You have led an interesting life! 🥳🥳✌️✌️

[@Nanalyze]

The fact that I didn't completely eff things up is quite astounding, but there's still some time left ;) Joe P.

[@PremiumCollector.]

You're the man, Joe!

[@Nanalyze]

Thank you for the support!

[@BigSources]

Isn't advanced AI just the death of start-ups? Even with non-AI companies it's like this: More data than start-ups--> easier to train AI on said data--> better product/service. If you have low users, customers or really any kind of product data as a start-up, the AI advantage for product or service development is totally against you. No matter if it's logistics, service or r&d. AI will be able to improve all these sectors within a company with enough relevant data, which big companies have way more of than any start-up and since that data is unique to every company, it's not even necessarily like the AI models they use would be helpful to a start-up. Of course if you have a unique and useful product, you can overcome those factors, but it'll definitly be way more difficult to break through in any big markets now more than ever, that's my semi-educated guess atleast.

[@Nanalyze]

That's a very good assessment. We've always said this about insurtech startups.

[@fendicoathelpmehidegun5021]

Nice background Joe!

[@Nanalyze]

Thank you! I am fortunate to frequent some very nice places. Joe P.

[@mariyork]

Arigato gozaimasu always! 😊

[@Nanalyze]

We appreciate the support!

[@azulsimmons1040]

This is a tough area to invest at this stage. Big, safe tech seems most straightforward, but which smaller players to hop on with are harder to find.

[@Nanalyze]

Yes, big safe tech will probably come out ahead because they have the capital to invest, they can make mistakes, and they can kick the tires and acquire only what begins to show real traction.

[@antonystringfellow5152]

I don't so much think that generative AI will avoid the Gartner hype cycle, just that it won't be too relevant. The reason being that the field is advancing at a far faster pace than any previous technology. So fast in fact, that the reality will soon catch up to the hype. I think the current narrative regarding the hype, real though the hype is, is providing a great opportunity for those wanting to invest in the area - especially the profitable semiconductor companies. There's a lot of talk about these stocks being in a bubble, like the tech bubble of the 90s, and a lot of people actually believe this! This is creating a great opportunity for those of us who know better. If you're one of these people, just be patient, keep your eye on the news, and make sure you're ready when the market for these stocks turns. Some new, much more capable models are due for release any day now, and at least one real biggy next year. A lot of people are going to be surprised - don't be one of them.

[@Nanalyze]

So this time it's different? Nobody can say at all what happens, but just remember the four most dangerous words in investing. Also, there are a myriad of factors that affect the market that can impact stock performance aside from whether AI is hyped or not. Unknown unknowns humble even the most astute Nostradum types. ;)

[@TimRobertsen]

Digital Creutzfeldt-Jakob could be interesting :p

[@Nanalyze]

It's called Model Autophagy Disorder or MAD

[@Walker956]

a veggie omelet with eggs in it? you didnt even comment on that. wtf

[@Nanalyze]

You want the robot to make a veggie omelet? 🤔

[@mk-ki3jc]

Joe is clearly on the right side of the dunning kruger curve.

[@Nanalyze]

Don't stroke his ego please. It will only get worse.

[@mk-ki3jc]

@@Nanalyze haha, at least it seems well deserved.

[@wizzyno1566]

Synthetic data isnt data. Shocking that anyone think it is.

[@Nanalyze]

Controversial topic this one ;)

[@factotums]

I've been using Gemini/Bard since it was released (I think) and have noted almost monthly how its performance has degraded, especially regarding the expanding class of information deemed 'sensitive'...ie; I asked for a history of maps of the Caribbean the other day and got the response 'Google does not provide information pertaining to upcoming elections' and that's hardly an outlier when talking about that automated response. Last year you could have long form discussions about Palestinian history and now you can't even ask for the weather in Gaza. Literally. The word 'gaza' will send you an automated response each time. I tried a workaround where if you type like an idiot with a bunch of grammatical and spelling errors 'wat wedthr in guava' you can sometimes get an answer. Also I wanted to say something about gene editing and the potential legal nightmares ahead for it. Remembering back in the 90s, when genetically modified foods were banned in several countries, and that was just agriculture. Now you are talking about introducing actual 'synthetic biology' into the ecosystem. I'm far from an ESG nut but I can see this becoming a central political topic once the normal world starts making memes about super-terrorists or even just genetically modified liberals running the ATF. All of that said, I've been trying to up my coding since AI became a buzzword for progress, and the integration of Gemini into Google Colab has been extremely impressive. The other day I asked it to give me a ten part lesson plan for Machine Learning using pop culture coding metaphors and it did not disappoint. Note: I am not sponsored by Google, in fact I think they probably hate me over there. Cheers

[@Nanalyze]

"...you can't even ask for the weather in Gaza." Don't know whether to laugh or cry about that statement, but "wat wedthr in guava" made me lol. It's probably a given that nobody working on moderating what the thought police allow you to talk about has every been to the Middle East. Gene editing vs GMO is a great topic to discuss, and we covered that here (www.nanalyze.com/2017/06/gmo-vs-gene-editing-vs-genetic-engineering/). Gene edited people from any political persuasion are a concern. Genetically modified babies anyone? Covered here (www.nanalyze.com/2016/10/genetic-engineering-humans-germline/). Google had a major problem with activists and they seem to have moved past that now. Covered here (kzbin.info/www/bejne/bqemdZuem9R5mtk). Thank you for the informative and entertaining comment. (Takes content plugging sales cap off.) Joe P.

[@brandongillett2616]

A good career for your children when they grow up? Only if they are already in college pursueing a degree for this. AGI will not take 10-20 years it will be here in 3-4 years.

[@Nanalyze]

Lots of experts would disagree with that timeline while some would agree. So your suggestion is that kids should study what then? Our conclusions still hold unless we can be convinced otherwise.

[@brandongillett2616]

@@Nanalyze More experts agree with my time-line (3-4 years) than do yours (10-20 years or more). But the majority fall somewhere just past my time-line. The question, "What should children be studying?" Is not deep enough to recon with the magnitude of transformation we are about to undergo. A better question to ask is, "Why should children be studying?" I can't exactly say what the answer to that is, but I know the answer no longer involves ideas about getting a good career. Robotics and AI will outpace all of us in terms of capacity for productive output. Careers as we know them today will cease to exist.

[@Nanalyze]

@@brandongillett2616 We don't have a timeline. We simply look at what experts say. Do you have some source that says the majority of experts believe AGI will be here "somewhere just past 3-4 years?" Children should be studying so they have a more fulfilling life than just being entertained by social media. Until "careers cease to exist" it's probably a good idea to study STEM.

[@Nanalyze]

@brandongillett2616 This is useful: research.aimultiple.com/artificial-general-intelligence-singularity-timing/

[@MrSeadawg123]

What about giving AI access to the dark web?

[@Nanalyze]

Wonder what the ratio of "dark web" data vs "proper internet" data looks like? The former must just be a small chunk, no? One would think that training these LLMs involves having lots of controls (structure) on inputs so you can try to explain the behavior somewhat. The dark web is extremely unstructured by nature so perhaps not much of use to be found there. Not sure. Joe P.

[@MrSeadawg123]

@@Nanalyze The NSA has a program to make it completely searchable. The darkweb is like 96% of all the whole internet content. If i remember right.

[@Nanalyze]

@@MrSeadawg123 Great rabbit hole. Came across this on Reddit. "The "Deep Web" isn't synonymous with the "Dark Web". The Deep Web is just any website that isn't cataloged by search engines, such as Google and Bing. You may still be able to access it with your standard browser if you know the specific address. The Deep Web can also be any network that is connected to the internet but if password protected, like a university or a business's private network. It's not accessible directly, but still connected."

[@MrSeadawg123]

@@Nanalyze One of the things edward Snowden released. Was the program used by the NSA. To find users of it. And what they had been searching for. But the DEEP WEB has the majority of content.