I mean, if you look back 50 years, they look like low-hanging fruits, but at the time, they weren't, only with the benefit of hindsight do they look like low-hanging fruits.
Similarly people in 50 years will say we had all the low-hanging fruits available today in subject/area/topic X, although we don't see them, as we don't have hindsight yet.
Like, these things seem obvious in hindsight even now.
You already have the whole Internet of data. You already have GPUs. All you need to do is just use the GPUs to feed the scraped (and maybe not-so-legally downloaded) data into some artificial neural network and you'll get a rather intelligent AI! How could one possibly think otherwise?
Probably referring to having a sales team/person, rather than actual sales.
Hard to judge what is wrong/could be better as I couldn't gather what your actual project is. Maybe there are some obvious glaring mistakes on the landing page or something that scares people off?
Or it might just be trying to solve a problem that people don't actually have, or communicated poorly what problem it solves. I tend to see these being the most common reasons people don't even sign up.
Oh, the problem is there trust me. I am very close to both industries and I know for a fact that my service covers big holes. I'm just a sucker at converting.
I won't comment on the project itself, but more about your mindset. I don't think "converting" is your issue, but your mindset is getting in the way of yourself. You see it as a no-brainer yet don't understand why people aren't buying, but you're 100% convinced you're right and solves the right problem in the right way. But clearly, if it did, you wouldn't even have to do sales, take a look at anubis for a project that is doing the very same but sells itself.
So I guess my best tip to you is to look inwards, be humble, be honest and try to enter a "learning" mode instead of "telling/convinced" mode that you seem to currently be in.
The only relevant part I can find from that article is:
> One criticism we've seen of the FreePascal project in general concerns its documentation, although there is quite a lot of it: eight FPC manuals, and lengthy Lazarus docs in multiple languages. There is a paid-for tutorial e-book available, too.
The criticism is that there is too much documentation available? And they're long, and dare even to be available in multiple languages?
I'm incorrectly asking questions? You sound like a great author.
I was asking about why the documentation was bad, with one example. You could reply "No" but instead shared some word-salad?
You could have just replied "Someone told me the indexing and cross-referencing is poor, I agree/disagree with that because of X" so we could have a normal human conversation instead of you trying to lecture some random internet commentator on completely irrelevant logic. Just be human instead.
It seems perfectly fine, information-dense even which is even better. Seems a lot better than the typical one-long-landing-page-docs many languages have today. What exactly is the problem with the wiki that cannot be fixed and must be re-made from scratch?
The wiki is full of incomplete, obsolete, or otherwise not-so-useful articles. It suffers from typical "wiki as documentation" efforts, where instead of concentrated efforts from domain experts, you get a thousand half-baked opinions.
It has good stuff, but I'd wager the "bad stuff" outweighs it by a large margin.
The problem is optics. Pretend I'm a brand new user and I want to build GUI applications. I've heard of a language called Freepascal and of an IDE called Lazarus. I think they're connected but I don't know how. And what's Delphi?
Most IDEs and languages have a Documentation link. Which link do I use to start with?
FreePascal has lazarus docs and Lazarus has FreePascal docs?
As a new user I can slog through 4 different links of schtuff, I guess. (Disappointment and frustration lie ahead; broken old buggy software that doesn't match the documentation. New User doesn't know that yet.) Maybe I'll just look for Youtube videos, but my enthusiasm is draining.
I'm used to this:
https://go.dev/doc/
Everything linked from go.dev/doc works today and it's coherent. Everything.
As other responses mention, the docs as a rule are out of date and confusing.
We should just invent what we had for horses in cities but for dogs, a little poopy-collector dangling from their hips at all time, that catches it just like that guy in the park.
You don't know this, but instead assuming parent lives in some suburban area with lots of other cats. They could be living outside in the woods, 5km to any close settlement, with minimal side-effects of having a cat outside (besides the side-effect of having a few less rodents around).
But no, lets have a knee-jerk reaction to anyone who has an outside cat, without understanding any of the context.
Besides, many people put bells on their cats, and then they're unlikely to catch anything at all in the wilderness.
It seems that in the modern era of social media campaigns for everything there can hardly exist a perfectly chill, normal activity that someone hasn't somehow contrived into a type of moral and ethical sin. It's tedious, sad and ultimately, kneejerk stupid. I strongly doubt that the world's domesticated cat population is generally creating an ecological apocalypse and the studies I have seen around it are far from anything you could call solid. Either way, believe it or not, you can actually also use very practical solutions like bell collars to easily fix most of these situations. How about getting off your moral pedestal about such a silly "issue".
Maybe I've just had stupid cats, but they never managed to catch any birds, even when they were without bells. Plenty of mouse offerings though, but seems the bells help with that too.
Yeah, my opinion is a bit more pragmatic and attached to reality, where context, environment and your actions matter, not some "empirical" study done by universities.
Personally, I love my cats so I let them roam outside instead of keeping them inside like a prison. Then I also care about other animals so naturally they have a bell so they cannot (successfully) hunt other animals. But again, pragmatic approaches aren't for everyone, some people love books and/or data instead :)
The alternative to empiricism (science) is rationalism (wish-casting), not pragmatism (least harm).
I often let my dog off-leash. Weighing the risks & rewards, I pragmatically choose to break the law, knowing full well that I'm in the wrong, not some special case. I eat the tickets and social scorn without complaint. My dog has pretty good recall and is super gentle (esp w/ kids). But the big bad govt (and other parents) didn't write the laws with my special pooch in mind.
You're confident your cat doesn't harm birds. Terrific. It's still wrong, in the general case. So take your lumps.
A (huge) point in your favor is that 2/3rd of (domesticated) cats are feral. So keeping cats indoors in order to better protect birds seems quixotic.
In these parts, owners keep their cats indoors to protect them. Recently, my SO's cat escaped her "catio" and was swiftly caught by a coyote. (A neighbor saw it happen. Horrifying.) Maybe your locale doesn't have coyotes.
Edit: Another exception (that I can think of) is farm/barn cats. Pretty much a necessity. Alas, coyotes. And probably hawks.
Could be worth considering that outdoor cats in the US may actually be a positive because so many of our country's natural predators of rodents and birds have been wiped out.
True, this is why I always look around when my dog poops, so I can evaluate if I should pick it up or not. Usually I leave it so people like astura can get mad when they step on it.
> I would be impressed if a sophomore in high school proposed it
That sounds good enough for a start, considering you can massively parallelize the AI co-scientist workflow, compared to the timescale and physical scale it would take to do the same thing with human high school sophomores.
And every now and then, you get something exciting and really beneficial coming from even inexperienced people, so if you can increase the frequency of that, that sounds good too.
We don't need an army of high school sophomores, unless they are in the lab pipetting. The expensive part of drug discovery is not the ideation phase, it is the time and labor spent running experiments and synthesizing analogues.
As discussed elsewhere, Deepmind are also working on extending Alphafold to simulate biochemical pathways and then looking to tackle whole-cell simulation. It's not quite pipetting, but this sort of AI scientist would likely be paired with the simulation environment (essentially as function calling), to allow for very rapid iteration of in-silico research.
It sounds like you're suggesting that we need machines that mass produce things like automated pipetting machines and the robots that glue those sorts of machines together.
Replacing a skilled technician is remarkably challenging. Often times, when you automate this, you just end up wasting a ton of resources rather than accelerating discovery. Often, simply integrating devices from several vendors (or even one vendor) takes months.
I've built microscopes intended to be installed inside workcells similar to what companies like Transcriptic built (https://www.transcriptic.com/). So my scope could be automated by the workcell automation components (robot arms, motors, conveyors, etc).
When I demo'd my scope (which is similar to a 3d printer, using low-cost steppers and other hobbyist-grade components) the CEO gave me feedback which was very educational. They couldn't build a system that used my style of components because a failure due to a component would bring the whole system down and require an expensive service call (along with expensive downtime for the user). Instead, their mech engineer would select extremely high quality components that had a very low probability of failure to minimize service calls and other expensive outages.
Unfortunately, the cost curve for reliability not pretty, to reduce mechanical failures to close to zero costs close to infinity dollars.
One of the reasons Google's book scanning was so scalable was their choice to build fairly simple, cheap, easy to maintain machines, and then build a lot of them, and train the scanning individuals to work with those machines quirks. Just like their clusters, they tolerate a much higher failure rate and build all sorts of engineering solutions where other groups would just buy 1 expensive device with a service contract.
This sounds like it could be centralised, a bit like the clouds in the IT world. A low failure rate of 1-3% is comparable to servers in a rack, but if you have thousands of them, then this is just a statistic and not a servicing issue. Several hyperscalers simply leave failed nodes where they are, it’s not worth the bother to service them!
Maybe the next startup idea is biochemistry as a service, centralised to a large lab facility with hundreds of each device, maintained by a dedicated team of on-site professionals.
None of the companies that proposed this concept have managed to demonstrate strong marketplace viability. A lot of discovery science remains extremely manual, artisinal, and vehemently opposed to automation.
> They couldn't build a system that used my style of components because a failure due to a component would bring the whole system down and require an expensive service call
Could they not make the scope easily replaceable by the user and just supply a couple of spares?
Just thinking of how cars are complex machines but a huge variety of parts could be replaced by someone willing to spend a couple of hours learning how.
That’s similar to how Google won in distributed systems. They used cheap PCs in shipping containers when everyone else was buying huge expensive SUN etc servers.
yes, and that's the reason I went to work at google: to get access to their distributed systems and use ML to scale up biology. I never was able to join Google Research and do the work I wanted (but DeepMind went ahead and solved protein structure prediction, so, the job got done anyway).
They really didn't solve it. AF works great for proteins that have a homologous protein with a crystal structure. It is absolutely useless for proteins with no published structure to use as a template - e.g. many of the undrugged cancer targets in existence.
@dekhn it is true (I also work in the field. I'm a software engineer who got a wet-lab PhD in biochemistry and work at a biotech doing oncology drug discovery)
There is a big range in both automation capabilities and prices.
We have a couple automation systems that are semi-custom - the robot can handle operation of highly specific, non-standard instruments that 99.9% of labs aren't running. Systems have to handle very accurate pipetting of small volumes (microliters), moving plates to different stations, heating, shaking, tracking barcodes, dispensing and racking fresh pipette tips, etc. Different protocols/experiments and workflows can require vastly different setups.
So pharmaceutical research is largely an engineering problem, of running experiments and synthesizing molecules as fast, cheap and accurate as possible ?
I wouldn't say it's an engineering problem. Biology and pharmacology are very complex with lots of curveballs, and each experiment is often different and not done enough to warrant full engineering-scale optimization (although this is sometimes the case!).
It also seems to be a financial problem of getting VC funds to run trials to appease regulators. Even if you’ve already seen results in a lab or other country.
We could have an alternative system where VC don’t need to appease regulators but must place X billion in escrow for compensation of any harm the medicine does to customers.
Regulator is not only there to protect the public, it also protects VC from responsibility
Regulations around clinical trials represent the floor of what's ethically permissible, not the ceiling. As in, these guidelines represent the absolute bare minimum required when performing drug trials to prevent gross ethical violations. Not sure what corners you think are ripe for cutting there.
> Regulations around clinical trials represent the floor of what's ethically permissible, not the ceiling.
Disagree. The US FDA especially is overcautious to the point of doing more harm than good - they'd rather ban hundreds of lifesaving drugs than allow one thalidomide to slip through.
Yeah that's not how anything works. Compounds are approved for use or not based on empirical evidence, thus the need for clinical trials. What's your level of exposure to the pharma industry?
> Compounds are approved for use or not based on empirical evidence, thus the need for clinical trials.
But off-label use is legal, so it's ok to use a drug that's safe but not proven effective (to the FDA's high standards) for that ailment... but only if it's been proven effective for some other random ailment. That makes no sense.
> What's your level of exposure to the pharma industry?
I strongly encourage you to take a half hour and have a look at what goes into preclinical testing and the phases of official trials. An understanding of the data gathered during this process should clear up some of your confusion around safety and efficacy of off-label uses, which parenthetically pharma companies are strictly regulated against encouraging in any way.
This is the general problem with nearly all of this era of generative AI and why the public dislike it so much.
It is trained on human prose; human prose is primarily a representation of ideas; it synthesizes ideas.
There are very few uses for a machine to create ideas. We have a wealth of ideas and people enjoy coming up with ideas. It’s a solution built for a problem that does not exist.
Won't this always be the case?
I mean, if you look back 50 years, they look like low-hanging fruits, but at the time, they weren't, only with the benefit of hindsight do they look like low-hanging fruits.
Similarly people in 50 years will say we had all the low-hanging fruits available today in subject/area/topic X, although we don't see them, as we don't have hindsight yet.