AI can already create photo-realistic images, and the old "look at the hands" rule doesn't really work on images generated with modern models.
There may be a few tells still, but those won't last long, and the moment someone can find a new pattern you can make that a negative prompt for new images to avoid repeating the same mistake.
I think we are already there, and it seems like we aren't because many people are using free low-quality models with a low number of steps because its more accessible.
That's a comprehensive guide. If anyone wants a similar introduction, with interactive exercises to try while they study this is also a good resource: https://www.codecademy.com/learn/learn-r
The AI is multiple programs working together, and they already pass math problems on to a data analyst specialist. There's also an option to use a WolframAlpha plugin to handle math problems.
The reason it didn't have math from the start was that it was a solved problem on computers decades ago, and they are specifically demonstrating advances in language capabilities.
Machines can handle math, language, graphics, and motor coordination already. A unified interface to coordinate all of those isn't finished, but gluing together different programs isn't a significant engineering problem.
> The AI is multiple programs working together, and they already pass math problems on to a data analyst specialist. There's also an option to use a WolframAlpha plugin to handle math problems.
is quality of this system good enough to qualify for AGI?..
I guess we will know it when we see it. Its like saying computer graphics got so good that we have holodeck now. We dont have holodeck yet. We don't have AGI yet.
It's a shame that Gemini is so far behind ChatGPT. Gemini Advanced failed softball questions when I've tried it, but GPT works almost every time even when I push the limits.
Google wants to replace the default voice assistant with Gemini, I hope they can make up the gap and also add natural voice responses too.
You tried Gemini 1.5 or just 1.0? I got an invite to try 1.5 Pro which they said is supposed to be equivalent to 1.0 Ultra I think?
1.0 Ultra completely sucked but when I tried 1.5 it is actually quite close to GPT4.
It can handle most things as well as ChatGPT 4 and in some cases actually does not get stuck like GPT does.
I'd love to hear other peoples thoughts on Gemini 1.0 vs 1.5? Are you guys seeing the same thing?
I have developed a personal benchmark of 10 questions that resemble common tasks I'd like an AI to do (write some code, translate a PNG with text into usable content and then do operations on it, Work with a simple excel sheet and a few other tasks that are somewhat similar).
I recommend everyone else who is serious about evaluating these LLMs think of a series of things they feel an "AI" should be able to do and then prepare a series of questions. That way you have a common reference so you can quickly see any advancement (or lack of advancement)
GPT-4 kinda handles 7 of the 10. I say kinda because it also gets hung up on the 7th task(reading a game price chart PNG with an odd number of columns and boxes) depending on how you ask: They have improved over the last year slowly and steadily to reach this point.
>a personal benchmark of 10 questions that resemble common tasks
That is an idea worth expanding on. Someone should develop a "standard" public list of 100 (or more) questions/tasks against which any AI version can be tested to see what the program's current "score" is (although some scoring might have to assign a subjective evaluation when pass/fail isn't clear).
Thats what a benchmark is, and they're all gamed by everyone training models, even if they don't intend to, because the benchmarks are in the training data.
The advantage of a personal set of questions is that you might be able to keep it out of the training set, if you don't publish it anywhere, and if you make sure cloud-accessed model providers aren't logging the conversations.
Gemini 1.0 Pro < Gemini 1.5 Pro < Gemini 1.0 Ultra < GPT-4V
GPT-4V is still the king. But Google's latest widely available offering (1.5 Pro) is close, if benchmarks indicate capability (questionable). Gemini's writing is evidently better, and vastly more so its context window.
Its nice to have some more potentially viable competition. Gemini has better OCR capabilities but its computation abilities seem to fall short....so I have it do the work with the OCR and then move the remainder of the work to GPT4 :)
> Anecdotally, I've seen this a lot with adults with undiagnosed ADHD (including myself until 26). Caffeine can be a (very shitty) alternative to proper stimulant medication that people don't realise they actually need because of a medical issue. So the end up drinking 4-5 coffees a day instead.
ADHD medication is among the hardest to get in the first place, and to maintain in many places.
I was diagnosed as a child, and still couldn't get regular medication until 3 years ago (I'm 37 now). Doctors will try to push off-label treatments, often SSRIs, which do nothing for ADHD. They will push people to try ineffective talk therapy and describe actual medication as "only a last resort".
My cousin can't get it because his insurance says you only have ADHD if you were diagnosed as a child, and he spent his childhood in another country where the roads were barely maintained.
ADHD meds reduce the chances someone will abuse drugs according to research, but many docs will cancel prescriptions if you are honest about cannabis use.
In the wake of the opiate scandal new restrictions also place limits on how much a pharmacy can dispense, even if everyone has a valid prescription. That's part of the reason for the medication shortages in some areas.
It's been a miracle medication for me, but if you don't have money and luck it can be next to impossible to find someone to help you. Those that do often require thousands in fees.
To add: At least for methylphenidate, there are several studies showing neuronal long term improvements, "brain normalisation", in people with ADHD. That's rare in psychopharmacology...
I do understand hesitation about immediate release amphetamine, but the retarded formulations are also quite hard to abuse. If you do more than prescribed, you won't get high, you will have a bad time. "Getting high", requires a sudden onset, rush, and sharp peak, which retarded meds won't induce.
Denying people with ADHD to try stims, is really, really unethical considering our knowledge on their effectiveness treating the condition.
True. That's a whole nother can of worms. You often only get something like 30 days worth of pills per prescription before you have to visit your doctor again. For someone with poor time and tasks management, those 2-3 hours lost (and adding the "resources" spent on mental overhead/stress/anxiety), are actually a significant burden.
I don't see how that's related at all. Apple chargers were always just dumb USB-A bricks, they didn't care what device you were charging with them or what kind of port it had. I don't remember when was the last time any phone came with a charger with an integrated cable instead of just USB-A/C output.
There may be a few tells still, but those won't last long, and the moment someone can find a new pattern you can make that a negative prompt for new images to avoid repeating the same mistake.
I think we are already there, and it seems like we aren't because many people are using free low-quality models with a low number of steps because its more accessible.