AI May Kill Jobs Plus It Can Kill Bambi, Koalas, and Whales
March 8, 2024
This essay is the work of a dumb dinobaby. No smart software required.Amid the AI hype is little mention of a huge problem.
As Nature’s Kate Crawford reports, “Generative AI’s Environmental Costs Are Soaring—and Mostly Secret.” Besides draining us of fresh water, AI data centers also consume immense amounts of energy. We learn:
“One assessment suggests that ChatGPT, the chatbot created by OpenAI in San Francisco, California, is already consuming the energy of 33,000 homes. It’s estimated that a search driven by generative AI uses four to five times the energy of a conventional web search. Within years, large AI systems are likely to need as much energy as entire nations.”
Even OpenAI’s head Sam Altman admits this is not sustainable, but he has a solution in mind. Is he pursuing more efficient models, or perhaps redesigning data centers? Nope. Altman’s hopes are pinned on nuclear fusion. But that technology has been “right around the corner” for the last 50 years. We need solutions now, not in 2050 or later. Sadly, it is unlikely AI companies will make the effort to find and enact those solutions unless forced to. The article notes a piece of legislation, the Artificial Intelligence Environmental Impacts Act of 2024, has finally been introduced in the Senate. But in the unlikely event the bill makes it through the House, it may be too feeble to make a real difference. Crawford considers:
“To truly address the environmental impacts of AI requires a multifaceted approach including the AI industry, researchers and legislators. In industry, sustainable practices should be imperative, and should include measuring and publicly reporting energy and water use; prioritizing the development of energy-efficient hardware, algorithms, and data centers; and using only renewable energy. Regular environmental audits by independent bodies would support transparency and adherence to standards. Researchers could optimize neural network architectures for sustainability and collaborate with social and environmental scientists to guide technical designs towards greater ecological sustainability. Finally, legislators should offer both carrots and sticks. At the outset, they could set benchmarks for energy and water use, incentivize the adoption of renewable energy and mandate comprehensive environmental reporting and impact assessments. The Artificial Intelligence Environmental Impacts Act is a start, but much more will be needed — and the clock is ticking.”
Tick. Tock. Need a dead dolphin? Use a ChatGPT-type system.
Cynthia Murrell, March 8, 2024
Engineering Trust: Will Weaponized Data Patch the Social Fabric?
March 7, 2024
This essay is the work of a dumb dinobaby. No smart software required.
Trust is a popular word. Google wants me to trust the company. Yeah, I will jump right on that. Politicians want me to trust their attestations that citizen interest are important. I worked in Washington, DC, for too long. Nope, I just have too much first-hand exposure to the way “things work.” What about my bank? It wants me to trust it. But isn’t the institution the subject of a a couple of government investigations? Oh, not important. And what about the images I see when I walk gingerly between the guard rails. I trust them right? Ho ho ho.
In our post-Covid, pre-US national election, the word “trust” is carrying quite a bit of freight. Whom to I trust? Not too many people. What about good old Socrates who was an Athenian when Greece was not yet a collection of ferocious football teams and sun seekers. As you may recall, he trusted fellow residents of Athens. He end up dead from either a lousy snack bar meal and beverage, or his friends did him in.
One of his alleged precepts in his pre-artificial intelligence worlds was:
“We cannot live better than in seeking to become better.” — Socrates
Got it, Soc.
Thanks MSFT Copilot and provider of PC “moments.” Good enough.
I read “Exclusive: Public Trust in AI Is Sinking across the Board.” Then I thought about Socrates being convicted for corruption of youth. See. Education does not bring unlimited benefits. Apparently Socrates asked annoying questions which open him to charges of impiety. (Side note: Hey, Socrates, go with the flow. Just pray to the carved mythical beast, okay?)
A loss of public trust? Who knew? I thought it was common courtesy, a desire to discuss and compromise, not whip out a weapon and shoot, bludgeon, or stab someone to death. In the case of Haiti, a twist is that a victim is bound and then barbequed in a steel drum. Cute and to me a variation of stacking seven tires in a pile dousing them with gasoline, inserting a person, and igniting the combo. I noted a variation in the Ukraine. Elderly women make cookies laced with poison and provide them to special operation fighters. Subtle and effective due to troop attrition I hear. Should I trust US Girl Scout cookies? No thanks.
What’s interesting about the write up is that it provides statistics to back up this brilliant and innovative insight about modern life is its focus on artificial intelligence. Let me pluck several examples from the dot point filled write up:
- “Globally, trust in AI companies has dropped to 53%, down from 61% five years ago.”
- “Trust in AI is low across political lines. Democrats trust in AI companies is 38%, independents are at 25% and Republicans at 24%.”
- “Eight years ago, technology was the leading industry in trust in 90% of the countries Edelman studies. Today, it is the most trusted in only half of countries.”
AI is trendy; crunchy click bait is highly desirable even for an estimable survivor of Silicon Valley style news reporting.
Let me offer several observations which may either be troubling or typical outputs from a dinobaby working in an underground computer facility:
- Close knit groups are more likely to have some concept of trust. The exception, of course, is the behavior of the Hatfields and McCoys
- Outsiders are viewed with suspicion. Often for now reason, a newcomer becomes the default bad entity
- In my lifetime, I have watched institutions take actions which erode trust on a consistent basis.
Net net: Old news. AI is not new. Hyperbole and click obsession are factors which illustrate the erosion of social cohesion. Get used to it.
Stephen E Arnold, March 7, 2024
AI and Warfare: Gaza Allegedly Experiences AI-Enabled Drone Attacks
March 7, 2024
This essay is the work of a dumb dinobaby. No smart software required.
We have officially crossed a line. DeepNewz reveals: “AI-Enabled Military Tech and Indian-Made Hermes 900 Drones Deployed in Gaza.” It this what they mean by “helpful AI”? We cannot say we are surprised. The extremely brief write-up tells us:
“Reports indicate that Israel has deployed AI-enabled military technology in Gaza, marking the first known combat use of such technology. Additionally, Indian-made Hermes 900 drones, produced in collaboration between Adani‘s company and Elbit Systems, are set to join the Israeli army’s fleet of unmanned aerial vehicles. This development has sparked fears about the implications of autonomous weapons in warfare and the role of Indian manufacturing in the conflict in Gaza. Human rights activists and defense analysts are particularly worried about the potential for increased civilian casualties and the further escalation of the conflict.”
On a minor but poetic note, a disclaimer states the post was written with ChatGPT. Strap in, fellow humans. We are just at the beginning of a long and peculiar ride. How are those assorted government committees doing with their AI policy planning?
Cynthia Murrell, March 7, 2024
Poohbahs Poohbahing: Just Obvious Poohbahing
March 6, 2024
This essay is the work of a dumb dinobaby. No smart software required.
We’re already feeling the effects of AI technology in deepfake videos and soundbites and generative text. While our present circumstances are our the beginning of AI technology, so-called experts are already claiming AI has gone bananas. The Verge, a popular Silicon Valley news outlet, released a new podcast episode where they declare that, “The AIs Are Officially Out Of Control.”
AI generated images and text aren’t 100% accurate. AI images are prone to include extra limbs, false representations of people, and even entirely miss the prompt. AI generative text is about as accurate as a Wikipedia article, so you need to double check and edit the response. Unfortunately AI are only as smart as the datasets that program them. AIs have been called “racist”and “sexist” due to limited data. Google Gemini also has gone too far on diversity and inclusion returning images that aren’t historically accurate when asked to deliver.
The podcast panelists made an obvious point when the pundits said that Google’s results qualities have declined. Bad SEO, crappy content, and paid results pollute search. They claim that the best results Google returns are coming from Reddit posts. Reddit is a catch-all online forum that Google recently negotiated deal with to use its content to train AI. That’s a great idea, especially when Reddit is going public on the stock market.
The problem is that Reddit is full of trolls who do things for %*^ and giggles. While Reddit is a brilliant source of information because it is created by real people, the bad actors will train the AI-chatbots to be “racist” and “sexist” like previous iterations. The worst incident involves ethnically diverse Nazis:
“Google has apologized for what it describes as “inaccuracies in some historical image generation depictions” with its Gemini AI tool, saying its attempts at creating a “wide range” of results missed the mark. The statement follows criticism that it depicted specific white figures (like the US Founding Fathers) or groups like Nazi-era German soldiers as people of color, possibly as an overcorrection to long-standing racial bias problems in AI.”
I am not sure which is the problem: Uninformed generalizations, flawed AI technology capable of zapping billions in a few hours, or minimum viable products are the equivalent of a blue jay fouling up a sparrow’s nest. Chirp. Chirp. Chirp.
Whitney Grace, March 6, 2024
Just One Big Google Zircon Gemstone for March 5, 2024
March 5, 2024
This essay is the work of a dumb dinobaby. No smart software required.
I have a folder stuffed with Google gems for the week of February 26 to March 1, 12023. I have a write up capturing more Australians stranded by following Google Maps’s representation of a territory, Google’s getting tangled in another publisher lawsuit, Google figuring out how to deliver better search even when the user’s network connection sucks, Google’s firing 43 unionized contractors while in the midst of a legal action, and more.
The brilliant and very nice wizard adds, “Yes, we have created a thing which looks valuable, but it is laboratory-generated. And it is gem and a deeply flawed one, not something we can use to sell advertising yet”. Thanks, MSFT Copilot Bing thing. Good enough and I liked the unasked for ethnic nuance.
But there is just one story: Google nuked billions in market value and created the meme of the week by making many images the heart and soul of diversity. Pundits wanted one half of the Sundar and Prabhakar comedy show yanked off the stage. Check out Stratechery’s view of Google management’s grasp of leading the company in a positive manner in Gemini and Google’s Culture. The screw up was so bad that even the world’s favorite expert in aircraft refurbishment and modern gas-filled airships spoke up. (Yep, that’s the estimable Sergey Brin!)
In the aftermath of a brilliant PR move, CNBC ran a story yesterday that summed up the February 26 to March 1 Google experience. The title was “Google Co-Founder Sergey Brin Says in Rare Public Appearance That Company ‘Definitely Messed Up’ Gemini Image Launch.” What an incisive comment from one of the father of “clever” methods of determining relevance. The article includes this brilliant analysis:
He also commented on the flawed launch last month of Google’s image generator, which the company pulled after users discovered historical inaccuracies and questionable responses. “We definitely messed up on the image generation,” Brin said on Saturday. “I think it was mostly due to just not thorough testing. It definitely, for good reasons, upset a lot of people.”
That’s the Google “gem.” Amazing.
Stephen E Arnold, March 5, 2024
Synthetic Data: From Science Fiction to Functional Circumscription
March 4, 2024
This essay is the work of a dumb humanoid. No smart software required.
Synthetic data are information produced by algorithms, not by real-world events. It’s created using real-world data and numerical recipes. The appeal is that it is easier than collecting real life information, cheaper than dealing with data from real life, and faster than fooling around with surveys, monitoring devices, and law suits. In theory, synthetic data is one promising way of skirting the expense of getting humans involved.
“What Is [a] Synthetic Sample – And Is It All It’s Cracked Up to Be?” tackles the subject of a synthetic sample, a topic which is one slice of the synthetic data universe. The article seeks “to uncover the truth behind artificially created qualitative and quantitative market research data.” I am going to avoid the question, “Is synthetic data useful?” because the answer is, “Yes.” Bean counters and those looking to find a way out of the pickle barrel filled with expensive brine are going to chase after the magic of algorithms producing data to do some machine learning magic.
In certain situations, fake flowers are super. Other times, the faux blooms are just creepy. Thanks, MSFT Copilot Bing thing. Good enough.
Are synthetic data better than real world data? The answer from my vantage point is, “It depends.” Fancy math can prove that for some use cases, synthetic data are “good enough”; that is, the data produce results close enough to what a “real” data set provides. Therefore, just use synthetic data. But for other applications, synthetic data might throw some sand in the well-oiled marketing collateral describing the wonders of synthetic data. (Some university research labs are quite skilled in PR speak, but the reality of their methods may not line up with the PowerPoints used to raise venture capital.)
This essay discusses a research project to figure out if a synthetic sample works or in my lingo if the synthetic sample is good enough. The idea is that as long as the synthetic data is within a specified error range, the synthetic sample can be used and may produce “reliable” or useful results. (At least one hopes this is the case.)
I want to focus on one portion of the cited article and invite you to read the complete Kantar explanation.
Here’s the passage which snagged my attention:
… right now, synthetic sample currently has biases, lacks variation and nuance in both qual and quant analysis. On its own, as it stands, it’s just not good enough to use as a supplement for human sample. And there are other issues to consider. For instance, it matters what subject is being discussed. General political orientation could be easy for a large language model (LLM), but the trial of a new product is hard. And fundamentally, it will always be sensitive to its training data – something entirely new that is not part of its training will be off-limits. And the nature of questioning matters – a highly ’specific’ question that might require proprietary data or modelling (e.g., volume or revenue for a particular product in response to a price change) might elicit a poor-quality response, while a response to a general attitude or broad trend might be more acceptable.
These sentences present several thorny problems is academic speak. Let’s look at them in the vernacular of rural Kentucky where I live.
First, we have the issue of bias. Training data can be unintentionally or intentionally biased. Sample radical trucker posts on Telegram, and use those messages to train a model like Reor. That output is going to express views that some people might find unpalatable. Therefore, building a synthetic data recipe which includes this type of Telegram content is going to be oriented toward truck driver views. That’s good and bad.
Second, a synthetic sample may require mixing data from a “real” sample. That’s a common sense approach which reduces some costs. But will the outputs be good enough. The question then becomes, “Good enough for what applications?” Big, general questions about how a topic is presented might be close enough for horseshoes. Other topics like those focusing on dealing with a specific technical issue might warrant more caution or outright avoidance of synthetic data. Do you want your child or wife to die because the synthetic data about a treatment regimen was close enough for horseshoes. But in today’s medical structure, that may be what the future holds.
Third, many years ago, one of the early “smart” software companies was Autonomy, founded by Mike Lynch. In the 1990s, Bayesian methods were known but some — believe it or not — were classified and, thus, not widely known. Autonomy packed up some smart software in the Autonomy black box. Users of this system learned that the smart software had to be retrained because new terms and novel ideas not in the original training set were not findable by the neuro linguistic program’s engine. Yikes, retraining requires human content curation of data sets, time to retrain the system, and the expense of redeploying the brains of the black boxes. Clients did not like this and some, to be frank, did not understand why a product did not work like an MG sports car. Synthetic data has to be trained to “know” about new terms and avid the “certain blindness” probability based systems possess.
Fourth, the topic of “proprietary data modeling” means big bucks. The idea behind synthetic data is that it is cheaper. Building proprietary training data and keeping it current is expensive. Is it better? Yeah, maybe. Is it faster? Probably not when humans are doing the curation, cleaning, verifying, and training.
The write up states:
But it’s likely that blended models (human supplemented by synthetic sample) will become more common as LLMs get even more powerful – especially as models are finetuned on proprietary datasets.
Net net: Synthetic data warrants monitoring. Some may want to invest in synthetic data set companies like Kantar, for instance. I am a dinobaby, and I like the old-fashioned Stone Age approach to data. The fancy math embodies sufficient risk for me. Why increase risk? Remember my reference to a dead loved one? That type of risk.
Stephen E Arnold, March 4, 2023
Bad News Delivered via Math
March 1, 2024
This essay is the work of a dumb humanoid. No smart software required.
I am not going to kid myself. Few people will read “Hallucination is Inevitable: An Innate Limitation of Large Language Models” with their morning donut and cold brew coffee. Even fewer will believe what the three amigos of smart software at the National University of Singapore explain in their ArXiv paper. Hard on the heels of Sam AI-Man’s ChatGPT mastering Spanglish, the financial payoffs are just too massive to pay much attention to wonky outputs from smart software. Hey, use these methods in Excel and exclaim, “This works really great.” I would suggest that the AI buggy drivers slow the Kremser down.
The killer corollary. Source: Hallucination is Inevitable: An Innate Limitation of Large Language Models.
The paper explains that large language models will be reliably incorrect. The paper includes some fancy and not so fancy math to make this assertion clear. Here’s what the authors present as their plain English explanation. (Hold on. I will give the dinobaby translation in a moment.)
Hallucination has been widely recognized to be a significant drawback for large language models (LLMs). There have been many works that attempt to reduce the extent of hallucination. These efforts have mostly been empirical so far, which cannot answer the fundamental question whether it can be completely eliminated. In this paper, we formalize the problem and show that it is impossible to eliminate hallucination in LLMs. Specifically, we define a formal world where hallucination is defined as inconsistencies between a computable LLM and a computable ground truth function. By employing results from learning theory, we show that LLMs cannot learn all of the computable functions and will therefore always hallucinate. Since the formal world is a part of the real world which is much more complicated, hallucinations are also inevitable for real world LLMs. Furthermore, for real world LLMs constrained by provable time complexity, we describe the hallucination-prone tasks and empirically validate our claims. Finally, using the formal world framework, we discuss the possible mechanisms and efficacies of existing hallucination mitigators as well as the practical implications on the safe deployment of LLMs.
Here’s my take:
- The map is not the territory. LLMs are a map. The territory is the human utterances. One is small and striving. The territory is what is.
- Fixing the problem requires some as yet worked out fancier math. When will that happen? Probably never because of no set can contain itself as an element.
- “Good enough” may indeed by acceptable for some applications, just not “all” applications. Because “all” is a slippery fish when it comes to models and training data. Are you really sure you have accounted for all errors, variables, and data? Yes is easy to say; it is probably tough to deliver.
Net net: The bad news is that smart software is now the next big thing. Math is not of too much interest, which is a bit of a problem in my opinion.
Stephen E Arnold, March 1, 2024
Student Surveillance: It Is a Thing
March 1, 2024
This essay is the work of a dumb dinobaby. No smart software required.
Once mobile phones were designed with cameras, all technology was equipped with one. Installing cameras and recording devices is SOP now, but facial recognition technology will soon become as common unless privacy advocates have their way. Students at the University of Waterloo were upset to learn that vending machines on their campus were programmed with the controversial technology. The Kitchener explores how the scandal started in: “ ‘Facial Recognition’ Error Message On Vending Machine Sparks Concern At University Of Waterloo.”
A series of smart vending machines decorated with M&M graphics and dispense candy were located throughout the Waterloo campus. They raised privacy concerns when a student noticed an error message about the facial recognition application on one machine. The machines were then removed from campus. Until they were removed, word spread quickly and students covered a hole believed to hold a camera.
Students believed that vending machines didn’t need to have facial recognition applications. They also wondered if there were more places on campus where they were being monitored with similar technology.
The vending machines are owned by MARS, an international candy company, and manufactured by Invenda. The MARS company didn’t respond to queries but Invenda shared more information about the facial recognition application:
“Invenda also did not respond to CTV’s requests for comment but told Stanley in an email ‘the demographic detection software integrated into the smart vending machine operates entirely locally.’ ‘It does not engage in storage, communication, or transmission of any imagery or personally identifiable information,’ it continued.
According to Invenda’s website, the Smart Vending Machines can detect the presence of a person, their estimated age and gender. The website said the ‘software conducts local processing of digital image maps derived from the USB optical sensor in real-time, without storing such data on permanent memory mediums or transmitting it over the Internet to the Cloud.’”
Invenda also said the software is compliant with the European Union privacy General Data Protection Regulation but that doesn’t mean it is legal in Canada. The University of Waterloo has asked that the vending machines be removed from campus.
Net net: Cameras will proliferate and have smart software. Just a reminder.
Whitney Grace, March 1, 2024
Surprise! Smart Software and Medical Outputs May Kill You
February 29, 2024
This essay is the work of a dumb humanoid. No smart software required.
Have you been inhaling AI hype today? Exhale slowly, then read “Generating Medical Errors: GenAI and Erroneous Medical References,” produced by the esteemed university with a reputation for shaking the AI cucarachas and singing loudly “Ai, Ai, Yi.” The write up is an output of the non-plagiarizing professionals in the Human Centered Artificial Intelligence unit.
The researchers report states:
…Large language models used widely for medical assessments cannot back up claims.
Here’s what the HAI blog post states:
we develop an approach to verify how well LLMs are able to cite medical references and whether these references actually support the claims generated by the models. The short answer: poorly. For the most advanced model (GPT-4 with retrieval augmented generation), 30% of individual statements are unsupported and nearly half of its responses are not fully supported.
Okay, poorly. The disconnect is that the output sounds good, but the information is distorted, off base, or possibly inappropriate.
What I found interesting is a stack ranking of widely used AI “systems.” Here’s the chart from the HAI article:
The least “poor” are the Sam AI-Man systems. In the middle is the Anthropic outfit. Bringing up the rear is the French “small” LLM Mistral system. And guess which system is dead last in this Stanford report?
Give up?
The Google. And not just the Google. The laggard is the Gemini system which was Bard, a smart software which rolled out after the Softies caught the Google by surprise about 14 months ago. Last in URL validity, last in statement level support, and last in response level support.
The good news is that most research studies are non reproducible or, like the former president of Stanford’s work, fabricated. As a result, I think these assertions will be easy for an art history major working in Google’s PR confection machine will bat them away like annoying flies in Canberra, Australia.
But last from researchers at the estimable institution where Google, Snorkel and other wonderful services were invented? That’s a surprise like the medical information which might have unexpected consequences for Aunt Mille or Uncle Fred.
Stephen E Arnold, February 29, 2024
A Fresh Outcry Against Deepfakes
February 29, 2024
This essay is the work of a dumb humanoid. No smart software required.
There is no surer way in 2024 to incur public wrath than to wrong Taylor Swift. The Daily Mail reports, “’Deepfakes Are a Huge Threat to Society’: More than 400 Experts and Celebrities Sign Open Letter Demanding Tougher Laws Against AI-Generated Videos—Weeks After Taylor Swift Became a Victim.” Will the super duper mega star’s clout spur overdue action? Deepfake porn has been an under-acknowledged problem for years. Naturally, the generative AI boom has made it much worse: Between 2022 and 2023, we learn, such content has increased by more than 400 percent. Of course, the vast majority of victims are women and girls. Celebrities are especially popular targets. Reporter Nikki Main writes:
“‘Deepfakes are a huge threat to human society and are already causing growing harm to individuals, communities, and the functioning of democracy,’ said Andrew Critch, AI Researcher at UC Berkeley in the Department of Electrical Engineering and Computer Science and lead author on the letter. … The letter, titled ‘Disrupting the Deepfake Supply Chain,’ calls for a blanket ban on deepfake technology, and demands that lawmakers fully criminalize deepfake child pornography and establish criminal penalties for anyone who knowingly creates or shares such content. Signees also demanded that software developers and distributors be held accountable for anyone using their audio and visual products to create deepfakes.”
Penalties alone will not curb the problem. Critch and company make this suggestion:
“The letter encouraged media companies, software developers, and device manufacturers to work together to create authentication methods like adding tamper-proof digital seals and cryptographic signature techniques to verify the content is real.”
Sadly, media and tech companies are unlikely to invest in such measures unless adequate penalties are put in place. Will legislators finally act? Answer: They are having meetings. That’s an act.
Cynthia Murrell, March 29, 2024
-
- Subscribe to Beyond Search
Feature archive
News archive
-
