I recently subscribed to Perplexity Pro and prior to this release, was already strongly considering discontinuing ChatGPT Premium.
When I first subscribed to ChatGPT Premium late last year, the natural language understanding superiority was amazing. Now the benchmark advances, low latency voice chat, Sora, etc. are all really cool too.
But my work and day-to-day usage really rely on accurately sourced/cited information. I need a way to comb through an ungodly amount of medical/scientific literature to form/refine hypotheses. I want to figure out how to hard reset my car's navigation system without clicking through several SEO-optimized pages littered with ads. I need to quickly confirm scientific facts, some obscure, with citations and without hallucinations. From speaking with my friends in other industries (e.g. finance, law, construction engineering), this is their major use case too.
I really tried to use ChatGPT Premium's Bing powered search. I also tried several of the top rated GPTs - Scholar AI, Consensus, etc.. It was barely workable. It seems like with this update, the focus was elsewhere. Unless I specify explicitly in the prompt, it doesn't search the web and provide citations. Yeah, the benchmark performance and parameter counts keep impressively increasing, but how do I trust that those improvements are preventing hallucinations when nothing is cited?
I wonder if the business relationship between Microsoft and OpenAI is limiting their ability to really compete in AI driven search. Guessing Microsoft doesn't want to disrupt their multi-billion dollar search business. Maybe the same reason search within Gemini feels very lacking (I tried Gemini Advanced/Ultra too).
I have zero brand loyalty. If anybody has a better suggestion, I will switch immediately after testing.
In the same situation as you. Genomics data mining with validated LMM responses would be a godsend. Even more so when combined with rapid conversational interactions.
We are not far from the models asking themselves questions. Recurrence will be ignition = first draft AGI. Strap in everybody.
Here's another vote for 3Blue1Brown's series on linear algebra [1]. Spending a few hours on this series will easily save you dozens of hours when going through a comprehensive LA textbook.
This is actually co-written/edited by Joe Blitzstein, who teaches Harvard's Stat110. What originally sold me on taking a look is this snippet from the foreword:
"This book is primarily a teaching tool, and thus we are not very concerned with proofs unless they promote understanding...There are no Statistics or linear algebra prerequisites."
Richard McElreath's content is a breath of fresh air for anyone who's struggled with stats.
I read both editions of his textbook and will be revisiting this new lecture material soon. I highly recommend you check out his book/course if you've been frustrated with trying to learn stats and want to practically understand things without hundreds of pages of proofs.
This book had a big influence on me. I highly recommend it. I read both the first and second editions.
It is very clearly written, full of practical examples (with code), and doesn't assume heavy math knowledge.
This section from the preface sums up the intended audience:
"The principle audience is researchers in the natural and social sciences, whether new PhD students or seasoned professionals, who have had a basic course on regression but nevertheless remain uneasy about statistical modeling. This audience accepts that there is something vaguely wrong about typical statistical practice in the early twenty-first century, dominated as it is by p-values and a confusing menagerie of testing procedures. They see alternative methods in journals and books. But these people are not sure where to go to learn about these methods."
Seems easy enough to play around with, and need not be strictly numbers either, as long as rank is defined on your fields (they are sortable...total ordering? I'm rusty on my terminology). Basically sort X, take the variance of the rank of Y, Z, etc. Much in the same way you would compute multi variable correlation.
Have either of you tried Pleese? I believe they're mostly selling to pizzerias now [1] and not supermarkets. They aren't able to scale up mass-production quickly because they use a different approach/ingredients - "proprietary blend of bean and potato proteins" [2].
I eat regular cheese and have tried several vegan brands, this seems pretty close to me. But I'm no cheese connoisseur.
Both the AstraZeneca and J&J vaccines use an adenovirus to deliver DNA instead of mRNA wrapped in lipid (like Moderna & Pfizer).
Everywhere I read about the J&J vaccine, I see something like "the DNA vaccine doesn't alter your DNA". Can somebody please clear this up?
As far as I understand, the mRNA just stays in the cytoplasm of the cell and gets used up by the ribosome to create spike proteins. The adenovirus vector used in the J&J (and other vaccines) injects DNA in the cell's nucleus, which seems at odds with the widely circulated "it doesn't change your DNA" statement.
Do people make this claim because the cell displaying spike proteins is basically always eliminated by CD8 killer T cells?
I'm a bioengineer. Everyone that gives an answer mostly ignores epigenetics and the fact that RNA can permanently alter the _shape_ (or conformation) of how your DNA is tightly wound up. For example, RNA can direct methylation of DNA and alter histones, which can lead to transgenerational epigenetic effects on gene expression and phenotype [1]. The fact is, molecular cell biology is incredibly complex and the models we have are just that, models. Saying that these mRNA strands don't affect the genome long-term may be correct, but this is an educated guess based on theoretical models. There's a reason why FDA approvals traditionally took 10+ years, we normally verify our educated guesses empirically.
It almost certainly won't have long-term affects, but it may not be trivial to identify if mRNA vaccines have been altering epigenetics.
Follow-up question: Do these effects (which are very unlikely) differ between vaccine and the virus itself? Or maybe: Is the epigenetic risk higher/different?
Everything in life has potential epigenetic effects! Not sure what you are trying to communicate with your comment but it sounds like you are saying people should avoid the mRNA vax but then you also say “almost certainly no long term effects.” Your thesis seems to be that 10 years is enough time to know for sure that they are safe. Why is 10 years the right amount of time? Why not 50 years or 5? In other words what’s your model for relative risk/reward and why is it better than what is being done in terms of public health outcomes?
You seem to be putting a lot of words in OPs mouth. I didn't see OP making any sort of suggestion about whether to avoid the vaccine or not. It looked like they were simply explaining some possible outcomes of the vaccine that are unknown.
OP also didn't say 10 years is enough time to know the long-term effects of these vaccines, just that it's traditionally been the minimum amount of time needed for some other drugs.
50 would certainly cover all life stages of humans. I assume you could be sure with shorter.
I have an aviation, biochem, and skydiving background. My rule is for aviation: "if it hasn't been out five years you're a test jumper."
Humans are way more complex than airplanes. I personally wouldn't take the mRNA vaccine because of this rule. Coupled with being unable to sue or get help from the government I think people IN LOW RISK groups have been way too enthusiastic to sign up.
I'd be happy to take the vaccine if I was in a higher risk group and I'll be happy to take the vaccine in a year or two, but right now I just don't think it's right decision for someone like myself.
Given my risk is very low I'm not too worried about COVID, but I am a little worried (perhaps wrongly) about the risk of finding out about some long-term side effect from these vaccines a few months down the road. I suffer from long-term side effects from another drug I took in the past, and at the time I was told there was no risk of long-term side effects and that it was safe to take. Only recently has the labeling been updated to reflect the discovery that permanent side effects can occur in some cases and for me it's too late, but I learnt my lesson to allow others to be the guinea pig for new drugs wherever possible.
It's really quite alarming how little we know about the body, espically considering the certainty of some "experts" about how extremely low the risk of adverse effects are from newly approved vaccines. I'm aware of a number of drugs which are approved and frequently perscribed which we don't even understand the mechanism of action for -- accutane, for example. Of course in this situation, we do know the mechanism of action, but it would still be wrong to assume we know the full surface area of possible side effects which could occur because our model of the human body is so basic.
I'm happy for someone to explain why I'm wrong on this. I'm obviously not an expert, just an average guy trying to assess the relative risk of two very unlikely events.
There is no reason to expect that side effects from the vaccine are not present or actually dramatically higher from the live virus. An infection by the virus, even if asymptomatic, will likely introduce way more alien genetic material and viral proteins into your body than the non-reproducing vaccine ever would. So one should trust the FDA panel of experts on risk unless there's strong evidence pointing otherwise.
> I think people IN LOW RISK groups have been way too enthusiastic to sign up.
About that...
> In December, we asked, “What percentage of people who have been infected by the coronavirus needed to be hospitalized?”
> The correct answer is not precisely known, but it is highly likely to be between 1% and 5% according to the best available estimates, and it is unlikely to be much higher or lower. We discuss the data and logic behind this conclusion in the appendix.
> Less than one in five U.S. adults (18%) give a correct answer of between 1 and 5%. Many adults (35%) say that at least half of infected people need hospitalization.
From what I can deduce using CDC data my chance of hospitalization is 0.5 to 2%. Assuming linear relationship of obesity in the population and assuming the same risk at the top of my cohort to the bottom.
Or maybe people are just terrible at judging acute risks? This isn't unique to covid - ask them about flying on an airliner or living next to a nuclear power plant and you would get some equally comical numbers. At any rate risk of hospitalization/death isn't the complete picture since some of those young people are ostensibly doing it to protect the people around them.
So let’s say hypothetically that white people were low risk relative people of color. You would have white people not get vaccinated because the consequences for their group, alone, might be slightly better?
Does that not seem a bit ... immoral?
I assume you eschew all other medical advances that are less than 50 years old? Would you eschew remdesivir? Sorry for the questions but thinking such as yours intrigues me and I want to grasp the logic behind it. Why not avoid all new technology for 50 years? Getting vaccinated seems to me like the logical and moral thing to do, but maybe I’m overlooking something.
Would this be an advantage for the Novavax vaccine as it doesn't do anything to hijack cellular machinery to create the spike protein, it just (as I understand it) has a bunch of pre-made spike proteins.
> Adenoviruses -- even as they occur in nature -- just do not have the capacity to alter DNA. Unlike retroviruses such as HIV or lentiviruses, wild-type adenoviruses do not carry the enzymatic machinery necessary for integration into the host cell's DNA. That's exactly what makes them good vaccine platforms for infectious diseases, according to Coughlan.
> And, engineered adenoviruses used in vaccines have been further crippled by deleting chunks of their genome so that they cannot replicate, further increasing their safety.
This sounds a bit like a technicality. The DNA makes it into the cell nucleus and is used by the cell machinery to make proteins. The changes aren't carried over after cell division, but lots of cells in your body last your whole life (nerves, brain cells, eye cells, important stuff).
Loss of neurons and cardiac muscle cells is permanent. Emergency medical personnel are usually taught "time is brain" and "time is heart" for this reason.
I think the argument hinges on the technicality that it's not splicing itself into the host genome, so no chance of it becoming a retrovirus or something like that (in the event that the cell's lineage is not extinguished by the immune system).
I'm not a genetic engineer (what a time to be alive, eh?), but I'm pretty sure an adenovirus that did permanently modify cell DNA would be more like CRISPR, including the risks that entails (such as the risk of incorrectly splicing the host genome and potentially creating a precancerous mutation)
I want to be perfectly clear that I didn't bring this up to be alarmist. Jesse Gelsinger's death shed a lot of light on the risks involved with adenoviruses [1]. Those lessons have been carried forward.
>> An autopsy and subsequent studies indicated that his death was caused by a fulminant immune reaction (with high serum levels of the cytokines interleukin-6 and interleukin-10) to the adenoviral vector.
>> The data suggested that the high dose of Ad [adenoviral] vector, delivered by infusion directly to the liver, quickly saturated available receptors ... within that organ and then spilled into the circulatory and other organ systems including the bone marrow, thus inducing the systemic immune response.
He was injected with >3 × 10^13 viruses [2]. The typical J&J dose contain: low-dose (5x10^10 viral particles) or high-dose (1x10^11 viral particles) [3].
Parent specifically stated DNA viruses. From your article:
> Most of these viral genes come from retroviruses, RNA viruses that insert DNA copies of their own genes into our genomes when they infect cells. HHV-6 is unique because it is the only known human DNA herpesvirus that integrates into the human genome and can be routinely inherited.
When I first subscribed to ChatGPT Premium late last year, the natural language understanding superiority was amazing. Now the benchmark advances, low latency voice chat, Sora, etc. are all really cool too.
But my work and day-to-day usage really rely on accurately sourced/cited information. I need a way to comb through an ungodly amount of medical/scientific literature to form/refine hypotheses. I want to figure out how to hard reset my car's navigation system without clicking through several SEO-optimized pages littered with ads. I need to quickly confirm scientific facts, some obscure, with citations and without hallucinations. From speaking with my friends in other industries (e.g. finance, law, construction engineering), this is their major use case too.
I really tried to use ChatGPT Premium's Bing powered search. I also tried several of the top rated GPTs - Scholar AI, Consensus, etc.. It was barely workable. It seems like with this update, the focus was elsewhere. Unless I specify explicitly in the prompt, it doesn't search the web and provide citations. Yeah, the benchmark performance and parameter counts keep impressively increasing, but how do I trust that those improvements are preventing hallucinations when nothing is cited?
I wonder if the business relationship between Microsoft and OpenAI is limiting their ability to really compete in AI driven search. Guessing Microsoft doesn't want to disrupt their multi-billion dollar search business. Maybe the same reason search within Gemini feels very lacking (I tried Gemini Advanced/Ultra too).
I have zero brand loyalty. If anybody has a better suggestion, I will switch immediately after testing.