I am also a scientist and had the same conclusion. I just use it to summarize papers, occasionally write boilerplate, and sometimes do some google search primitives if its an easy question.
I'd be surprised if they even breed a dog at this point. Most George Church companies go up in extraordinary flames and mismanagement. The constant VC funds his companies raise have probably destroyed the biotech ecosystem for the last decade +.
Private Equity is the effect not the cause. We need them to create efficiency because of the shenanigans that the AMA guild did in limiting doctor supply. Just allow people to take an exam to get credentialed, we'd have foreign doctors flown in by the hundreds of thousands and care would be as cheap as it is in India.
The notion of "stem cells" are somewhat of a misnomer as modern biotechnology is showing us. There are an orchestration of probably 1000s of different cell types which can vary in their cell states. We don't have enough knowledge of how to use these various cell types to influence complex self-organization processes like development. But in some cases there are clever fixes without understanding the underlying process.
China is going to be way ahead of us in biological treatments because they are willing to actually remove the red tape and in fact encourage scientists to try these sorts of experimental treatments. Meanwhile, we have a dinosaur FDA, a bureaucratic academia second to only Europeans.
We don't accept that? Our civil liberties have been infringed upon by neoliberals and oligarchs for the last 100+ years, starting with Lincoln with income tax and FDR with social security into the mess we have today.
Because we in america tend to value individual liberties in a way that others don't. And yes of course LPR is mass surveillance. Cars should not even have to be registered.
Car culture in general is an abomination to civil rights. You are tracked by the government, forced to shell out money to predatory insurance companies, and can basically be illegally searched and seized at any point on the roadway. I hope that chinese e-bikes/e-motos will get good enough to where I can use them as my primary form of transport.
I just get the same videos recommended over and over. I liked it when the YT feed would recommend stuff new and different i might find interesting. Now its just hyperoptimized to get me to click on ads.
Definitely not. It’s more like either winning a 1B B2B contract or going bankrupt. Like I said above you don’t want to scale up taking high variance shots because it will reduce variance and converge towards the mean
To be clear, I meant the opposite. I meant that if you take a high variance swing and you lose, you have capped downside, so you can rebuild and recover hard long term. Conversely if you win you have a huge head start and can deploy the capital quickly.
Of course you DONT want to repeatedly take -EV bets. That would reduce variance and converge you to the mean which is negative by the central limit theorem.
It's mostly SV grifters who shoved LLMs up our asses. They then get in cahoots with boomers in the government to create policies and "investment schemes" that inflate their stock in a ponzi-like fashion and regulate competition.
Why do you think Trump has some no-name crypto firm, or why Thiel has Vance as his whipping boy, and Elon spend a fortune trying to get Trump to win? This is a multiparty thing, as most politicians are heavily bought and paid for.
Who would buy OpenAI's spy device? I think a lot of public discourse and backlash about the greedy, anticompetitive, and exploitative practices of the silicon valley elite have gone mainstream and will hopefully course correct the industry in time.
Boomers in government would be clueless on how to properly regulate and create correct incentives. Hell, that is still a bold ask for tech and economist geniuses with the best of intentions.
Because its an event contract with a defined upside/downside and time horizon. You know exactly what you stand to lose and gain and when. Makes it a valuable part of some intricate financial strategies.
Once the free money runs out, the AI cos may shift to making heavily verified code snippets with more direct language control. This will heavily simplify a lot of boilerplate instead of fairytales of some AGI coding wiz.
Exactly, I think perplexity had the right idea of where to go with AI (though obviously fumbled execution). Essentially creating more advanced primitives for information search and retrieval. So it can be great at things we have stored and need to perform second order operations on (writing boilerplate, summarizing text, retrieving information).
That is doubtful. Most of the content on people's feed was basically converging to the same paradigm anyways. Think the Mr. Beast "I gave someone 1 million dollars to try a potato" content.
Precisely. If companies would just focus on what it could be good at - deductive search, coding boilerplate with assistance, etc. then it would be a great tool. Instead you have dario, altman, and co. trying to pump stock and give us more spaghetti agents.
Great article and exactly why I use AI less and less. I basically find it to be rotting my brain towards the middle of the distribution. It's like all the nuance and critical thinking that actually goes into things gets stripped out.
Once a company perfects an agent that essentially performs condensed search and coding boilerplate making, that is probably where LLMs end for me. Perplexity and Claude are on the right track but not at all close.