2023 Impact Players [sic] per Chat GPT

gtchem05

Jolly Good Fellow
Messages
390
If you're looking for an accurate assessment of the impact players for 2023, you'll want to look elsewhere, but if you want to have an off-season laugh or two at the expense of Chat GPT, then you've come to the right place.

Here's the response I received when I asked the chatbot, "Who will be the impact players for georgia tech football in the fall of 2023?":

As the fall 2023 season approaches, Georgia Tech football fans are eager to see how their team will perform in the Atlantic Coast Conference. The Yellow Jackets have a challenging schedule ahead of them, facing off against some of the top teams in the nation. To have a successful season, Georgia Tech will need some of its players to step up and make a big impact on the field. Here are some of the potential impact players for Georgia Tech football fall 2023:

- Quarterback: Jake Smith. The sophomore signal-caller showed flashes of brilliance in his freshman year, throwing for 2,134 yards and 18 touchdowns with only six interceptions. He also ran for 467 yards and four scores, displaying his dual-threat ability. Smith has a strong arm and good mobility, but he needs to improve his accuracy and decision-making under pressure. If he can develop into a more consistent and confident leader, he could be one of the best quarterbacks in the ACC.

- Running back: Jahmyr Gibbs. The junior running back was a breakout star in his sophomore year, rushing for 1,256 yards and 14 touchdowns on 6.4 yards per carry. He also caught 32 passes for 312 yards and three touchdowns, proving to be a versatile weapon in the offense. Gibbs has a rare combination of speed, power and vision, and he can make defenders miss with his elusiveness and agility. He is the focal point of Georgia Tech's rushing attack, and he could challenge for the ACC rushing title in 2023.

- Wide receiver: Malachi Carter. The senior wide receiver is the most experienced and reliable target in Georgia Tech's passing game. He led the team in receptions (49), receiving yards (689) and receiving touchdowns (seven) in his junior year. He has good size (6'3", 205 lbs), hands and route-running skills, and he can make tough catches in traffic. He is Smith's go-to receiver on third downs and in the red zone, and he could have a career year in his final season.

- Defensive end: Kevin Harris. The senior defensive end is the leader and playmaker of Georgia Tech's defensive line. He had a stellar junior year, recording 12 sacks, 16 tackles for loss, four forced fumbles and two fumble recoveries. He has a quick first step, a powerful bull rush and a relentless motor, and he can disrupt opposing offenses with his pressure and penetration. He is one of the best pass rushers in the ACC, and he could earn All-American honors in 2023.

- Linebacker: Quez Jackson. The senior linebacker is the heart and soul of Georgia Tech's defense. He led the team in tackles (102), tackles for loss (13) and interceptions (three) in his junior year. He has great instincts, range and coverage skills, and he can make plays all over the field. He is the captain and communicator of the defense, and he sets the tone with his physicality and intensity. He is one of the best linebackers in the ACC, and he could be a finalist for the Butkus Award in 2023.
 

takethepoints

Helluva Engineer
Messages
6,142
I believe it. AI (the very term is a deception) is still in its infancy and would return just htis kind of flapdoodle. I expect that as language processzing improves, so will the quality of the results. No doubt you'll have a choice of whether you want an honest assessment base on the facts or something more positive or negative. The programs can already come close to that. Not, mind, that the results would be anything an actual human would trust.

I remember a Larry Niven story where Louis Wu's girlfriend had programmed two bots, a psychiatrist (Freud) and a physicist (Einstein) t6o help Louis with his work and his problems. The physicist gave Louis a weird answer and she went in to re-program it. She talked to the program, got it to fix its own code, and then tested it with questions suggested by her own version. I expect to see something like that in 40 - 50 years. Well … since I'm 77 I don't expect I will see it. Rats.
 

gtchem05

Jolly Good Fellow
Messages
390
They're pretty upfront that the training set of data used for ChatGPT is 2021 and prior.
That’s all fine and good, but even if you pretend the last two years never happened [which isn’t such a bad idea when it comes to GT football, now I think about it], I don’t seem to recall ever having a quarterback named Jake Smith at Georgia Tech, and I must have missed Kevin Harris’ 12-sack college football season. It seems like the chat bot is acting like someone with dementia in that it replaces its lack of memory with something factitious.
 

Jerry the Jacket

Helluva Engineer
Messages
1,959
Location
Chapin, SC
AI, much like self-driving vehicles, and renewable energy, are all much further down the road than portrayed in our media. I guess they are more interesting than the rest of the bull crap our modern media dishes out.
 

LongforDodd

LatinxBreakfastTacos
Messages
3,261
There is/was a Jake Smith who played QB at Blessed Trinity and who was recruited here for baseball. He is now in his fourth year at the AF Academy.
When Trevor Lawrence's regular season win streak was snapped, the last time he had lost a game was way back in high school and it was Blessed Trinity he lost to and Jake Smith was the QB.
 

iceeater1969

Helluva Engineer
Messages
9,764
There is/was a Jake Smith who played QB at Blessed Trinity and who was recruited here for baseball. He is now in his fourth year at the AF Academy.
When Trevor Lawrence's regular season win streak was snapped, the last time he had lost a game was way back in high school and it was Blessed Trinity he lost to and Jake Smith was the QB.
Chat LFD speaks!
 

bke1984

Helluva Engineer
Messages
3,599
They're pretty upfront that the training set of data used for ChatGPT is 2021 and prior.
It’s much worse than that though. This thing flat out makes stuff up that is no way rooted in fact or even rooted in data used to train the engine. The GPT folks call this “hallucination.” What worries me is that its confidence in how it presents answers that appear reasonably accurate but are completely false will be very dangerous as people start to incorporate this tool into their daily decision making.

But hey, the fantasy world it creates where we got Gibbs back is one I would thoroughly enjoy.
 

RonJohn

Helluva Engineer
Messages
5,044
It’s much worse than that though. This thing flat out makes stuff up that is no way rooted in fact or even rooted in data used to train the engine. The GPT folks call this “hallucination.” What worries me is that its confidence in how it presents answers that appear reasonably accurate but are completely false will be very dangerous as people start to incorporate this tool into their daily decision making.

But hey, the fantasy world it creates where we got Gibbs back is one I would thoroughly enjoy.
It is a tool that works pretty well for some things. As with all tools, and all new technologies, people should try to understand what the tool/technology can do and what it's limitations are. If you keep that in mind, ChatGPT is an excellent tool. Unfortunately, there are many other examples of technology that people jump straight into using in ways the tools are not ready for. Automatic braking and lane keep assist are excellent tools to help protect people in cars. Unfortunately, some people have ignored the limitations and believe they can go to sleep while driving such cars and still be safe. Face recognition is very good at identifying traits of faces and identifying people who reasonably match those traits. Unfortunately some police officers seem to believe that the output of facial recognition software is 100% accurate and have arrested innocent people with no other evidence.

If OpenAI was presenting ChatGPT as accurate, then I would have an issue with them. Since they do not, I blame users for ignoring the warnings that OpenAI posts about ChatGPT results. I haven't used it very much. When I have used it, it has been asking questions that I know the answer to in order to see if it can give the correct result. It has been very good at some things, and not so good at others. If I were to use it to actually find information that I don't already know the answer to, I would revert to my normal way of getting information. Be skeptical of any answer from anyone that you don't have full confidence in, and verify the information from everyone including those that you have full confidence in.
 

bke1984

Helluva Engineer
Messages
3,599
It is a tool that works pretty well for some things. As with all tools, and all new technologies, people should try to understand what the tool/technology can do and what it's limitations are. If you keep that in mind, ChatGPT is an excellent tool. Unfortunately, there are many other examples of technology that people jump straight into using in ways the tools are not ready for. Automatic braking and lane keep assist are excellent tools to help protect people in cars. Unfortunately, some people have ignored the limitations and believe they can go to sleep while driving such cars and still be safe. Face recognition is very good at identifying traits of faces and identifying people who reasonably match those traits. Unfortunately some police officers seem to believe that the output of facial recognition software is 100% accurate and have arrested innocent people with no other evidence.

If OpenAI was presenting ChatGPT as accurate, then I would have an issue with them. Since they do not, I blame users for ignoring the warnings that OpenAI posts about ChatGPT results. I haven't used it very much. When I have used it, it has been asking questions that I know the answer to in order to see if it can give the correct result. It has been very good at some things, and not so good at others. If I were to use it to actually find information that I don't already know the answer to, I would revert to my normal way of getting information. Be skeptical of any answer from anyone that you don't have full confidence in, and verify the information from everyone including those that you have full confidence in.
I hear you, but you are clearly someone who scrutinizes things they are told - as am I. Not everyone is that way (I believe the vast majority of people are the opposite). Unfortunately there are some that are attempting to make money on this technology by building and marketing tools that suggest accuracy in the responses, despite being fully aware of the limitations you’ve cited. While the danger of these issues is much more subtle than that of a self driving car, it is very real. However, this subtlety means that everyone is welcoming this stuff with open arms without taking the time to think about how dangerous it is when everyone starts relying on false information.

The good news is that I’m certain this is the most GPT has discussed Tech football, and given the small relative size of our fan base the risk of widespread misinformation regarding our program is very limited (unless of course we start running the 3O again - when it was extremely common)!
 
Top