TheLatestNate (@TheLatestNate) header
TheLatestNate (@TheLatestNate) profile picture

Tweetin' for fun For the love of humanity! Where shall wisdom be found?

A series of tubes
Joined August 2018
Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)Photo by TheLatestNate (@TheLatestNate)
Seriously people. The time is now I stood back for awhile but came out of the woodwork once I saw how quickly things were going to deteriorate We need people in power who can meet this moment, across governments and businesses alike
Wake up!!! @azeem is calling all these hacks “AI’s Wuhan moment”. Marc Benioff called it a code red. And it is clear it is all going to get worse. Yet the AI companies insist on shipping obviously dangerous products. Meanwhile the U.S. government lacks the will and/or competence to do anything about it. We are on an express train to hell. Tell your congress people to temporarily ban AI agents, or brace for impact.
1
4
1
8
2,298
I need you guys to pay attention here because I found a pattern among these "one shot Ai, no code" scammers, who keep bragging about the games they "made". They all have something in common, I suspect they're all marketing accounts for TesanaAi, or Tripo, doesn't really matter, what matters is this is a common trick I remember in crypto that people would use to hype new NFTs and meme coins. Fake marketing accounts acting like what they produce or found is organic, there is maybe some truth behind them but for the most part, these aren't real people. In this example both accounts started posting some time this year, they also post the same thing every few days, until eventually it gets traction. Nothing is organic here, I also suspect many of them are connected in a sort of content farm like way. Please get the word out, because none of what you are seeing out there is real when it comes to this.
7
11
1
88
10,848
Increasingly worried this is the case and will once again be handled with short sighted profit seeking instead of actual soul searching and sustainable policy changes
The market is crashing because they can no longer deny that current LLMs past a certain point will be good enough in most legal purposes, and open weight models could potentially surpass frontier models because of the unrestricted freedom. Stopping Jenson won't stop what's inevitable, they only stopped a Mag7 company from endorsing the truth. LLMs are commodities and will be swapped out to serve the purposes needed for cheapest cost. Frontier models have no moat. The thing about building God, is its only profitable if they can control the gates of Heaven. If it's not profitable, there is no investment thesis that makes sense for the current datacenter buildout, and this runs down the line to every bottleneck compute company. This is a total selloff, despite these companies making record profits now, because the cut off will eventually be swifter than the growth they enjoyed. Fact is: Without AI the past three years, the market would have stagnated or become bearish. They were hoping it would save the West, to buy us enough time to bring back manufacturing and production that is desperately needed for a healthy, competitive society. The wars, the tarrifs. These were last ditch efforts to regain control. These didn't work either.
12
TheLatestNate retweeted
Ten classic works of philosophy that are under 200 pages
These are all relatively short works that introduce you to some of the greatest writers in literary history, and several can be finished in a single weekend: - Plato's Apology - Sophocles's Antigone - Voltaire's Candide - Goethe's The Sorrows of Young Werther - Tolstoy's The Death of Ivan Ilyich - Dostoevsky's White Nights - Kafka's The Metamorphosis - Hemingway's The Old Man and the Sea - Steinbeck's Of Mice and Men What would you add?
15
89
4
962
24,733
5
305
1
6,108
50,206
TheLatestNate retweeted
3
272
1
3,893
22,696
TheLatestNate retweeted
Kokiri Forest
15
313
2,475
19,630
TheLatestNate retweeted
I was diagnosed with stage 4 lung cancer in July 2024. Never smoked. Not even a ciggy on drunk nights out. Nobody knows the true cause, but radon is a leading candidate. If you're in a hot zone, check your radon levels. A radon removal system cost me about $1,000. Someone handy can do it for less. It's just piping under your house with a fan that exhausts it above the roofline, instead of letting it slowly trickle up through your home.
77
93
10
1,330
359,331
We're blessed to have Laika bring this story to a wider audience Go see Wildwood to get an idea of how excellent that will be And read this book if you haven't already!
Given the hate and discourse about modern fantasy fiction, I'll just say this is one of the best novels of recent history (and a beautiful example of elegant worldbuilding without lore dumps and faux-scientific rules): bookshop.org/a/88357/9781635…
40
Kaoru Yamada
3
55
1
433
4,650
Biggest issue with current social platforms. Rage and entertainment reign supreme You have to put active effort in to build networks and find channels that go against this Lots of wonderful people still post great things but they are increasingly buried in slop 🧵
It’s so sad to open twitter and just read nonsense. We built the greatest information-sharing technology in history, only to reward those who make people angrier instead of smarter.
1
14
You see what you engage with. Consistently reposting artists puts them in my feed You also see posts similar to your own. Making positive comments brings me positive posts And following the right people is key. Folks who share their favorite things and people instead of trends
1
6
All of this takes effort Don't feel like you need to get it done in a day or spend even more time on socials Spend less! Use it better. Save your favorite things. Share them with friends. Keep notes outside of these websites And never become beholden to any platform
4
TheLatestNate retweeted
I made EXACTLY this point 25 years ago in my first book, and still barely anyone understands how important it is. Props, as ever, to @ChombaBupe.
The universal approximation theorem states, any neural network (NN) with at least one hidden layer & enough nonlinear activation nodes in the hidden layers, can approximate any function but it doesn't say that that NN can learn any function given input-output pairs.
7
18
1
102
8,888
I want to believe 👽 There is potential for the mania to end without financial meltdown and funding re-routed to more beneficial causes Agree a bubble pop doesn't mean that AI is finished. It may even be the start of new innovations Important moment and a precarious time
Analysis of where we now stand With SoftBank cooked, Nvidia hurting, and the UAE probably feeling like they just overpaid, not to mention IPOs losing their appeal and bankers becoming very hesitant, I am just not seeing where the next giant wad of money to sustain expensive unprofitable businesses is coming from. That is going to leave a lot of companies bankrupt and others sold for parts. AI will rise again but many in this generation of companies are in deep trouble.
10
TheLatestNate retweeted
Nothing went wrong, the whole plan was to rush Anthropic to IPO, pull off a pump and dump, and suck up all available liquidity left in the market while opening the Emergency Exit door for all pre-IPO investors to cash out before it’s too late This also means OpenAI is doomed
Anthropic IPO shares were predominantly allocated only to existing investors and large AMs that will place them in their index strategies and ETFs and won’t lend shares in the market for short selling. 13% to retail investors. More than 50% on institutional orders got zero allocation but I have been told those have not been spotted buying what they didn’t get in the open market. From one side GS and MS made sure $ANT shares would not be flipped, but from another side I have also been told that they heavily supported the price today to make sure the close will be well above IPO price buffering heavy insiders selling at the beginning of trading While they and the company will claim this IPO has been a success, in reality $ANT will have poor organic demand and liquidity that will remain dependent on the bookrunners for the time being. Considering $ANT unlock schedule, its price will have a hard time to perform especially after the additional forced buying that will be triggered by index inclusion in ~15 days will be exhausted. Furthermore, the reason why both SP500 and Nasdaq are heavily down today signals that there wasn’t extra liquidity in the market to accommodate $ANT IPO that it might end up being the very last one of this AI bubble. Let’s also not forget that in little over a month Dems will take control of the Congress and the administration of key states like Texas with a strong mandate from voters to curb heavily the so far wild data centers buildout and financial speculation around that.
120
936
50
5,319
132,706
TheLatestNate retweeted
I love this perspecitve on evolution. Scientists have long pointed out that, since natural selection can only modify what already exists, organisms often have "inefficient" designs. Things an engineer never would have designed, like eye neurons wired backwards, or the panda's "thumb" being an elongated wrist bone. But another way to view the same phenomenon is, as @NikoMcCarty describes, natural selection painstakingly extracting every ounce of functionality out of limited hardware. We have a hard enough time designing molecules that can do one thing well, but natural selection frequently designs molecules with multiple functions! It reminds me of the ingenious design tricks early video game developers would employ to maximize what their games could do, despite weak systems with limited memory and slow data rates.
Evolution works in multifactorial ways, whereas human biologists tend to study things one dimension at a time. The tools we use in biology are reductionist, which means we only get a narrow view of how things work; our measurements are inherently limited in time or space. We sequence DNA and look at the letters, while ignoring (usually) the structure in which those letters fold up into 3D forms, which in turn determine how the letters are expressed in the cell. Of course, evolution doesn’t care about these distinctions. It can make use of structure or chemistry or size or any other feature. It can even "bake" multiple functions into the same molecule. There’s an enzyme called lactate dehydrogenase B, for example, that converts between lactate and pyruvate, but that same protein, when expressed in the eyeballs of ducks, instead acts as a crystallin protein and forms the lens. There are many similar examples of "dual" evolution in viruses. Hepatitis D has a tiny RNA genome with just one protein-coding gene. But the RNA itself is also catalytic; it folds up into a shape that then cleaves other RNA molecules to help the virus replicate itself. Evolution has optimized the same molecule to both carry information but also to have a particular structure. I’m often underwhelmed by the ability of bioengineering to find optimal solutions, and I think part of the reason is that we struggle to operate across these different dimensions. You might object and think that machine learning will help us engineer biology more easily, and that's true, but I’m skeptical even the best AI model could replicate the performance of evolution, at least so long as our measurement tools give us a narrow view, the data we provide to models will be similarly narrow. If we can’t even measure the interactions that matter for what we’re optimizing, etc etc.... I’m not the first to say this, of course. The great François Jacob wrote about evolution’s tinkering in a 1977 essay in Science, called “Evolution and Tinkering,” writing: "The action of natural selection has often been compared to that of an engineer. This, however, does not seem to be a suitable comparison. … It works like a tinkerer … The tinkerer, in contrast, always manages with odds and ends. What he ultimately produces is generally related to no special project, and it results from a series of contingent events, of all the opportunities he had to enrich his stock with leftovers."
7
21
3
405
15,931
TheLatestNate retweeted
She didn’t hold back.
64
1,506
160
7,742
440,114
TheLatestNate retweeted
The CEO of Sesame Street recently remarked they know they’re leaving millions on the table by not marketing an AI Elmo but they feel that AI toys break the parent/child bond by taking over the role of caretaker and given how actual AI toys act they made the right choice
Honey, bedtime! Did you brush your teeth and discuss the Rwandan genocide with Mr. Fluffy?
12
933
6
11,563
108,222
TheLatestNate retweeted
This makes my blood boil. Go watch Norm instead (Directed by @aronfromm). It’s a great reflection piece from Norm Macdonald about Sam, comedy, and mortality. Beautiful animated and directed. One of the best pieces of media i’ve seen this year. youtu.be/1kHWG9fGNQw?si=yV5M…
24
981
12
10,176
278,098
Yet another reason people will become entrenched against AI Why would anyone think this was the right thing to do? Is there really no better use of our time and resources?
16