Hitting the Center Field Wall, AI Suffers an Injury!
November 15, 2023
This essay is the work of a dumb, dinobaby humanoid. No smart software required.
At a reception at a government facility in Washington, DC, last week, one of the bright young sparks told me, “Every investment deal I see gets fund if it includes the words ‘artificial intelligence.’” I smiled and moved to another conversation. Wow, AI has infused the exciting world of a city built on the swampy marge of the Potomac River.
I think that the go-go era of smart software has reached a turning point. Venture firms and consultants may not have received the email with this news. However, my research team has, and the update contains information on two separate thrusts of the AI revolution.
The heroic athlete, supported by his publicist, makes a heroic effort to catch the long fly ball. Unfortunately our star runs into the wall, drops the ball, and suffers what may be a career-ending injury to his left hand. (It looks broken, doesn’t it?)Oh, well. Thanks, MSFT Bing. The perspective is weird and there is trash on the ground, but the image is good enough.
The first signal appears in “AI Companies Are Running Out of Training Data.” The notion that online information is infinite is a quaint one. But in the fever of moving to online, reality is less interesting that the euphoria of the next gold rush or the new Industrial Revolution. Futurism reports:
Data plays a central role, if not the central role, in the AI economy. Data is a model’s vital force, both in basic function and in quality; the more natural — as in, human-made — data that an AI system has to train on, the better that system becomes. Unfortunately for AI companies, though, it turns out that natural data is a finite resource — and if that tap runs dry, researchers warn they could be in for a serious reckoning.
The information or data in question is not the smog emitted by modern automobiles’ chip stuffed boxes. Nor is the data the streams of geographic information gathered by mobile phone systems. The high value data are those which matter; for example, in a stream of security information, which specific stock is moving because it is being manipulated by one of those bright young minds I met at the DC event.
The article “AI Companies Are Running Out of Training Data” adds:
But as data becomes increasingly valuable, it’ll certainly be interesting to see how many AI companies can actually compete for datasets — let alone how many institutions, or even individuals, will be willing to cough their data over to AI vacuums in the first place. But even then, there’s no guarantee that the data wells won’t ever run dry. As infinite as the internet seems, few things are actually endless.
The fix is synthetic or faked data; that is, fabricated data which appears to replicate real-life behavior. (Don’t you love it when Google predicts the weather or a smarty pants games the crypto market?)
The message is simple: Smart software has ground through the good stuff and may face its version of an existential crisis. That’s different from the rah rah one usually hears about AI.
The second item my team called to my attention appears in a news story called “OpenAI Pauses New ChatGPT Plus Subscriptions De to Surge in Demand.” I read the headline as saying, “Oh, my goodness, we don’t have the money or the capacity to handle more users requests.”
The article expresses the idea in this snappy 21st century way:
The decision to pause new ChatGPT signups follows a week where OpenAI services – including ChatGPT and the API – experienced a series of outages related to high-demand and DDoS attacks.
Okay, security and capacity.
What are the implications of these two unrelated stories:
- The run up to AI has been boosted with system operators ignoring copyright and picking low hanging fruit. The orchard is now looking thin. Apples grow on trees, just not quickly and over cultivation can ruin the once fertile soil. Think a digital Dust Bowl perhaps?
- The friction of servicing user requests is causing slow downs. Can the heat be dissipated? Absolutely but the fix requires money, more than high school science club management techniques, and common sense. Do AI companies exhibit common sense? Yeah, sure. Everyday.
- The lack of high-value or sort of good information is a bummer. Machines producing insights into the dark activities of bad actors and the thoughts of 12-year-olds are grinding along. However, the value of the information outputs seems to be lagging behind the marketers’ promises. One telling example is the outright failure of Israel’s smart software to have utility in identifying the intent of bad actors. My goodness, if any country has smart systems, it’s Israel. Based on events in the last couple of months, the flows of data produced what appears to be a failing grade.
If we take these two cited articles’ information at face value, one can make a case that the great AI revolution may be facing some headwinds. In a winner-take-all game like AI, there will be some Sad Sacks at those fancy Washington, DC receptions. Time to innovate and renovate perhaps?
Stephen E Arnold, November 15, 2023
The Risks of Smart Software in the Hands of Fullz Actors and Worse
November 7, 2023
This essay is the work of a dumb humanoid. No smart software required.
The ChatGPT and Sam AI-Man parade is getting more acts. I spotted some thumbs up from Satya Nadella about Sam AI-Man and his technology. The news service Techmeme provided me with dozens of links and enticing headlines about enterprise this and turbo that GPT. Those trumpets and tubas were pumping out the digital version of Funiculì, Funiculà.
I want to highlight one write up and point out an issue with smart software that appears to have been ignored, overlooked, or like the iceberg possibly that sank the RMS Titanic, was a heck of a lot more dangerous than Captain Edward Smith appreciated.
The crowd is thrilled with the new capabilities of smart software. Imagine automating mundane, mindless work. Over the oom-pah of the band, one can sense the excitement of the Next Big Thing getting Bigger and more Thingier. In the crowd, however, are real or nascent bad actors. They are really happy too. Imagine how easy it will be to automate processes designed to steal personal financial data or other chinks in humans’ armor!
The article is “How OpenAI Is Building a Path Toward AI Agents.” The main idea is that one can type instructions into Sam AI-Man’s GPT “system” and have smart software hook together discrete functions. These functions can then deliver an output requiring the actions of different services.
The write up approaches this announcement or marketing assertion with some prudence. The essay points out that “customer chatbots aren’t a new idea.” I agree. Connecting services has been one of the basic ideas of the use of software. Anyone who has used notched cards to retrieve items related to one another is going to understand the value of automation. And now, if the Sam AI-Man announcements are accurate that capability no longer requires old-fashioned learning the ropes.
The cited write up about building a path asserts:
Once you start enabling agents like the ones OpenAI pointed toward today, you start building the path toward sophisticated algorithms manipulating the stock market; highly personalized and effective phishing attacks; discrimination and privacy violations based on automations connected to facial recognition; and all the unintended (and currently unimaginable) consequences of infinite AIs colliding on the internet.
Fear, uncertainty, and doubt are staples of advanced technology. And the essay makes clear that the rule maker in chief is Sam AI-Man; to wit the essay says:
After the event, I asked Altman how he was thinking about agents in general. Which actions is OpenAI comfortable letting GPT-4 take on the internet today, and which does the company not want to touch? Altman’s answer is that, at least for now, the company wants to keep it simple. Clear, direct actions are OK; anything that involves high-level planning isn’t.
Let me introduce my observations about the Sam AI-Man innovations and the type of explanations about the PR and marketing event which has whipped up pundits, poohbahs, and Twitter experts (perhaps I should say X-spurts?)
First, the Sam AI-Man announcements strike me as making orchestration a service easy to use and widely available. Bad things won’t be allowed. But the core idea of what I call “orchestration” is where the parade is marching. I hear the refrain “Some think the world is made for fun and frolic.” But I don’t agree, I don’t agree. Because as advanced tools become widely available, the early adopters are not exclusively those who want to link a calendar to an email to a document about a meeting to talk about a new marketing initiative.
Second, the ability of Sam AI-Man to determine what’s in bounds and out of bounds is different from refereeing a pickleball game. Some of the players will be nation states with an adversarial view of the US of A. Furthermore, there are bad actors who have a knack for linking automated information to online extortion. These folks will be interested in cost cutting and efficiency. More problematic, some of these individuals will be more active in testing how orchestration can facilitate their human trafficking activities or drug sales.
Third, government entities and people like Sam AI-Man are, by definition, now in reactive mode. What I mean is that with the announcement and the chatter about automating the work required to create a snappy online article is not what a bad actor will do. Individuals will see opportunities to create new ways to exploit the cluelessness of employees, senior citizens, and young people. The cheerful announcements and the parade tunes cannot drown out the low frequency rumbles of excitement now rippling through the bad actor grapevines.
Net net: Crime propelled by orchestration is now officially a thing. The “regulations” of smart software, like the professionals who will have to deal with the downstream consequences of automation, are out of date. Am I worried? For me personally, no, I am not worried. For those who have to enforce the laws which govern a social construct? Yep, I have a bit of concern. Certainly more than those who are laughing and enjoying the parade.
Stephen E Arnold, November 7, 2023
Missing Signals: Are the Tools or Analysts at Fault?
November 7, 2023
This essay is the work of a dumb humanoid. No smart software required.
Returning from a trip to DC yesterday, I thought about “signals.” The pilot — a specialist in hit-the-runway-hard landings — used the word “signals” in his welcome-aboard speech. The word sparked two examples of missing signals. The first is the troubling kinetic activities in the Middle East. The second is the US Army reservist who went on a shooting rampage.
The intelligence analyst says, “I have tools. I have data. I have real time information. I have so many signals. Now which ones are important, accurate, and actionable?” Our intrepid professionals displays the reality of separating the signal from the noise. Scary, right? Time for a Starbuck’s visit.
I know zero about what software and tools, systems and informers, and analytics and smart software the intelligence operators in Israel relied upon. I know even less about what mechanisms were in place when Robert Card killed more than a dozen people.
The Center for Strategic and International Studies published “Experts React: Assessing the Israeli Intelligence and Potential Policy Failure.” The write up stated:
It is incredible that Hamas planned, procured, and financed the attacks of October 7, likely over the course of at least two years, without being detected by Israeli intelligence. The fact that it appears to have done so without U.S. detection is nothing short of astonishing. The attack was complex and expensive.
And one more passage:
The fact that Israeli intelligence, as well as the international intelligence community (specifically the Five Eyes intelligence-sharing network), missed millions of dollars’ worth of procurement, planning, and preparation activities by a known terrorist entity is extremely troubling.
Now let’s shift to the Lewiston Maine shooting. I had saved on my laptop “Six Missed Warning Signs Before the Maine Mass Shooting Explained.” The UK newspaper The Guardian reported:
The information about why, despite the glaring sequence of warning signs that should have prevented him from being able to possess a gun, he was still able to own over a dozen firearms, remains cloudy.
Those “signs” included punching a fellow officer in the US Army Reserve force, spending some time in a mental health facility, family members’ emitting “watch this fellow” statements, vibes about issues from his workplace, and the weapon activity.
On one hand, Israel had intelligence inputs from just about every imaginable high-value source from people and software. On the other hand, in a small town the only signal that was not emitted by Mr. Card was buying a billboard and posting a message saying, “Do not invite Mr. Card to a church social.”
As the plane droned at 1973 speeds toward the flyover state of Kentucky, I jotted down several thoughts. Like or not, here these ruminations are:
- Despite the baloney about identifying signals and determining which are important and which are not, existing systems and methods failed bigly. The proof? Dead people. Subsequent floundering.
- The mechanisms in place to deliver on point, significant information do not work. Perhaps it is the hustle bustle of everyday life? Perhaps it is that humans are not very good at figuring out what’s important and what’s unimportant. The proof? Dead people. Constant news releases about the next big thing in open source intelligence analysis. Get real. This stuff failed at the scale of SBF’s machinations.
- The uninformed pontifications of cyber security marketers, the bureaucratic chatter flowing from assorted government agencies, and the cloud of unknowing when the signals are as subtle as the foghorn on cruise ship with a passenger overboard. Hello, hello, the basic analysis processes don’t work. A WeWork investor’s thought processes were more on point than the output of reporting systems in use in Maine and Israel.
After the aircraft did the thump-and-bump landing, I was able to walk away. That’s more than I can say for the victims of analysis, investigation, and information processing methods in use where moose roam free and where intelware is crafted and sold like canned beans at TraderJoe’s.
Less baloney and more awareness that talking about advanced information methods is a heck of a lot easier than delivering actual signal analysis.
Stephen E Arnold, November 7, 2023
test
Bankrupting a City: Big Software, Complexity, and Human Shortcomings Does the Trick
September 15, 2023
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
I have noticed failures in a number of systems. I have no empirical data, just anecdotal observations. In the last few weeks, I have noticed glitches in a local hospital’s computer systems. There have been some fascinating cruise ship problems. And the airlines are flying the flag for system ineptitudes. I would be remiss if I did not mention news reports about “near misses” at airports. A popular food chain has suffered six recalls in a four or five weeks.
Most of these can be traced to software issues. Others are a hot mess combination of inexperienced staff and fouled up enterprise resource planning workflows. None of the issues were a result of smart software. To correct that oversight, let me mention the propensity of driverless automobiles to mis-identify emergency vehicles or possessing some indifference to side street traffic at major intersections.
“The information technology manager looks at the collapsing data center and asks, “Who is responsible for this issue?” No one answers. Those with any sense have adopted the van life, set up stalls to sell crafts at local art fairs, or accepted another job. Thanks, MidJourney. I guarantee your sliding down the gradient descent is accelerating.
What’s up?
My person view is that some people do not know how complex software works but depend on it despite that cloud of unknowing. Other people just trust the marketing people and buy what seems better, faster, and cheaper than an existing system which requires lots of money to keep chugging along.
Now we have an interesting case example that incorporates a number of management and technical issues. Birmingham, England is now bankrupt. The reason? The cost of a new system sucked up the cash. My hunch is that King Charles or some other kind soul will keep the city solvent. But the idea of city going broke because it could not manage a software project is illustrative of the future in my opinion.
“Largest Local Government Body in Europe Goes Under amid Oracle Disaster” reports:
Birmingham City Council, the largest local authority in Europe, has declared itself in financial distress after troubled Oracle project costs ballooned from £20 million to around £100 million ($125.5 million).
An extra £80 million would make little difference to an Apple, Google, or Microsoft. To a city in the UK, the cost is a bit of a problem.
Several observations:
- Large project management expertise does not deliver functional solutions. How is that air traffic control or IRS system enhancement going?
- Vendors rely on marketing to close deals, and then expect engineers to just make the system work. If something is incomplete or not yet coded, the failure rate may be anticipated, right? Nope, what’s anticipated in a scope change and billing more money.
- Government agencies are not known for smooth, efficient technical capabilities. Agencies are good at statements of work which require many interesting and often impossible features. The procurement attorneys cannot spot these issues, but those folks ride herd on the legal lingo. Result? Slips betwixt cup and lip.
Are the names of the companies involved important? Nope. The same situation exists when any enterprise software vendor wins a contract based on a wild and wooly statement of work, managed by individuals who are not particularly adept at keeping complex technical work on time and on target, and when big outfits let outfits sell via PowerPoints and demonstrations, not engineering realities.
Net net: More of these types of cases will be coming down the pike.
Stephen E Arnold, September 15, 2023
Generative AI: Not So Much a Tool But Something Quite Different
August 24, 2023
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
Thirty years ago I had an opportunity to do a somewhat peculiar job. I had written for a publisher in the UK a version of a report my team and I prepared about Japanese investments in its Fifth Generation Computer Revolution or some such government effort. A wealthy person who owned a medium-sized financial firm asked me if I would comment on a book called The Meaning of the Microcosm. “Sure,” I said.
This tiny, cute technology creature has just crawled from the ocean, and it is looking for lunch. Who knew that it could morph into a much larger and more disruptive beast? Thanks, MidJourney. No review committee for me this morning.
What I described was technology’s Darwinian behavior. I am not sure I was breaking new ground, but it seemed safe for me to point to how a technology survived. Therefore, I argued in a private report to this wealthy fellow, that if betting on a winner would make one rich. I tossed in an idea that I have thought about for many years; specifically, as technologies battle to “survive,” the technologies evolve and mutate. The angle I have commented about for many years is simple: Predicting how a technology mutates is a tricky business. Mutations can be tough to spot or just pop up. Change just says, “Hello, I am here.”
I thought about this “book commentary project” when I read “How ChatGPT Turned Generative AI into an Anything Tool.” The article makes a number of interesting observations. Here’s one I noted:
But perhaps inadvertently, these same changes let the successors to GPT3, like GPT3.5 and GPT4, be used as powerful, general-purpose information-processing tools—tools that aren’t dependent on the knowledge the AI model was originally trained on or the applications the model was trained for. This requires using the AI models in a completely different way—programming instead of chatting, new data instead of training. But it’s opening the way for AI to become general purpose rather than specialized, more of an “anything tool.”
I am not sure that “anything tool” is a phrase with traction, but it captures the idea of a technology that began as a sea creature, morphing, and then crawling out of the ocean looking for something to eat. The current hungry technology is smart software. Many people see the potential of combining repetitive processes with smart software in order to combine functions, reduce costs, or create alternatives to traditional methods of accomplishing a task. A good example is the use college students are making of the “writing” ability of free or low cost services like ChatGPT or You.com.
But more is coming. As I recall, in my discussion of the microcosm book, I made the point that Mr. Gilder’s point that small-scale systems and processes can have profound effects on larger systems and society as a whole. But a technology “innovation” like generative AI is simultaneously “small” and “large”. Perspective and point of view are important in software. Plus, the innovations of the transformer and the larger applications of generative AI to college essays illustrate the scaling impact.
What makes AI interesting for me at this time is that genetic / Darwinian change is occurring across the scale spectrum. On one hand, developers are working to create big applications; for instance, SaaS solutions that serve millions of users. On the other hand, shifting from large language models to smaller, more efficient methods of getting smart aim to reduce costs and speed the functioning of the plumbing.
The cited essay in Arstechnica is on the right track. However, the examples chosen are, it seems to me, ignoring the surprises the iterations of the technology will deliver. Is this good or bad? I have no opinion. What is important than wild and crazy ideas about control and regulation strike me as bureaucratic time wasting. It was millions a years ago to get out of the way of the hungry creature from the ocean of ones and zeros and try to figure out how to make catch the creature and have dinner, turn its body parts into jewelry which can be sold online, or processing the beastie into a heat-and-serve meal at Trader Joe’s.
My point is that the generative innovations do not comprise a “tool.” We’re looking at something different, semi-intelligent, and evolving with speed. Will it be let’s have lunch or one is lunch?
Stephen E Arnold, August 24, 2023
A Group without a Leader: Lost in the Digital Wilderness. No Signal, No Hope
August 10, 2023
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
I read a story in a business magazine which may not make executives at a certain company happy. In fact, some of these executives may be thinking about some form of digital retribution. The story concerns Google Maps, a Google product/service which I find is pretty much unusable. Keep in mind that I am a dinobaby and dinobaby talons can’t hit the miniature graphics which cover Google maps like my ninth-grade acne. (Yeah, ugly.)
A high technology company’s project team. The group is lost. No one has any idea what to do or which direction to take. Their manager told them, “Rely on the digital maps your colleagues made.” How is that working out for you? Thanks, MidJourney. You have the desperation look nailed.
“Google Maps Has Become an Eyesore. 5 Examples of How the App Has Lost Its Way” contains a list of issues the author who probably has more online experience than I do identifies five issues with the much-loved service. The “love” refers to the revenue generated from Google Maps, not the “love” or lack of it from users like me.
These range from icon acne to weird behaviors with the street name “feature.” I am not going to romp through the issues the article flags. I want to focus on two which are deal breakers for me. In fact, the digital map thing recently forced me to purchase a trucker’s edition of a printed road map to the United States.
For me, Google has made it difficult for me (probably not you, dear GenX reader) to find the street view. I quite like finding a location and then being able to look at the physical surroundings. How do I address this need now? I use Bing Maps.
The second issue that drives me crazy is the omission of businesses (intentionally or unintentionally) because the business does not advertise. I have written about the Cuba Libre Restaurant issue, and it bedevils me even today. I was standing in front of the bustling Washington, DC, restaurant, but my digital map service did not show it. Objectivity, they name is not Googzilla, I say.
Let me shift gears and offer my hypothesis why Google Maps is almost unusable for me.
Imagine a team responsible for a mapping product. There are a couple of people who have some tenure with the team. A couple have escaped from a more dysfunctional team; for example, a failed smart software project. Plus, there are two new hires with zero clue how or why they are working on maps. These individuals are experts in data center engineering and never leave the servers and, therefore, don’t know anything about maps, just wiring diagrams.
Okay, now the group sits around and someone says, “What are we supposed to do?” The most senior person who is totally occupied about getting on a hot team focused on killing another vendor’s AI effort, says, “Let’s just come up with some ideas and implement a couple.” The group mumbles, plays with their mobile devices, chats with the data center wizard about slow response on the internal messaging system, and look out the windows. One hard charger says, “Let’s make a list of ideas on the whiteboard, rank them, and do the top two or three.” More mumbles. A list is generated. The six team breaks into two groups and the employees retreat to the snack area to talk about implementing the functions. The work is agreed upon and the coding is dumped on the two network professionals. These individuals write the code, make sure it doesn’t kill anything, and emails it to the others on the team. No one looks at it but everyone says, “Fine.” Done.
This work procedure evidences:
- Zero guidance from an involved, motivated “manager”
- The mental attitude of the engineers
- The indifference of the individuals to the idea of delivering useful, quality features.
Now the author of the article about Google Maps knows nothing about this modern management approach to adding features at an unnamed high technology company.
That’s why I don’t rely on digital maps. The printed map works just fine. Plus I have to stop and read the map. None of the figure out a map driving or walking, which can lead to a collision with a smart, self driving automobile or an engineer looking for work at another company.
Stephen E Arnold, August 10, 2023
What Will Smart Software Change?
August 3, 2023
Note: Dinobaby here: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid. Services are now ejecting my cute little dinosaur gif. (´?_?`) Like my posts related to the Dark Web, the MidJourney art appears to offend someone’s sensibilities in the datasphere. If I were not 78, I might look into these interesting actions. But I am and I don’t really care.
Today (July 27, 2023) a person told me about “Photographs of People Making Books at the Collins Factory in 1960s Glasgow.” The write up is less compelling than the photographs. The online article features workers who:
- Organize products for shipping
- Setting type slugs with a hammer and chisel
- A person stitching book folios together
- A living artist making a plate
- A real person putting monotype back in a case.
I mention this because I have seen articles which suggest that smart software will not cause humans to lose their jobs. It took little time for publishers to cut staff and embrace modern production methods. It took less time for writers to generate a PDF and use an Amazon-type service to promote, sell, and distribute a book. Now smart software is allegedly poised to eliminate writers.
Will AI really create more work for humans?
The 1960s photos suggest that technology eliminates jobs in my opinion as it disrupts established work procedures and vaporizes norms which glue social constructs together. Anyone you know have the expertise to seat metal type with a hammer and chisel? I suppose I should have asked, “Does anyone near you scroll TikToks?”
Stephen E Arnold, August 3, 2023
The Authority of a Parent: In Question?
August 3, 2023
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
If we cannot scan the kids, let us scan the guardians. That is what the ESRB, digital identity firm Yoti, and kiddie marketing firm SuperAwesome are asking the Federal Trade Commission according to The Register‘s piece, “Watchdog Mulls Online Facial Age-Verification Tech—For Kids’ Parents.” The Children’s Online Privacy Protection Act (COPPA) requires websites and apps to make kids under 13 get a parent’s permission before they can harvest that sweet, early stage personal data. It is during the next step the petitioners would like to employ age-verification software on the grown-ups. As writer Jessica Lyons Hardcastle describes, the proposed process relies on several assumptions. She outlines the steps:
“1. First, a child visits a website and hits an age gate. The operator then asks the kid for their parent’s email, sends a note to the parent letting them know that they need to verify that they’re an adult for the child to proceed, and offers the facial-age scanning estimation as a possible verification method.
2. (Yes, let’s assume for a moment that the kid doesn’t do what every 10-year-old online does and lie about their age, or let’s assume the website or app has a way of recognizing it’s dealing with a kid, such as asking for some kind of ID.)
3. If the parent consents to having their face scanned, their system then takes a selfie and the software provides an age estimate.
4. If the age guesstimate indicates the parent is an adult, the kid can then proceed to the website. But if it determines they are not an adult, a couple of things happen.
5. If ‘there is some other uncertainty about whether the person is an adult’ then the person can choose an alternative verification method, such as a credit card, driver’s license, or social security number.
6. But if the method flat out decides they are not an adult, it’s a no go for access. We’re also going to assume here that the adult is actually the parent or legal guardian.”
Sure, why not? The tech works by converting one’s face into a set of numbers and feeding that to an AI that has been trained to assess age with those numbers. According to the ESRB, the actual facial scans are not saved for AI training, marketing, or any other purpose. But taking them, and their data-hungry partners, at their word is yet another assumption.
Cynthia Murrell, August 3, 2023
When Wizards Flail: The Mysteries of Smart Software
July 18, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[1]](http://arnoldit.com/wordpress/wp-content/uploads/2023/07/Vea4_thumb_thumb_thumb_thumb_thumb_t1-30.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[1]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
How about that smart software stuff? VCs are salivating. Whiz kids are emulating Sam AI-man. Users are hoping there is a job opening for a Wal-Mart greeter. But there is a hitch in the git along; specifically, some bright experts are not able to understand what smart software does to generate output. The cloud of unknowing is thick and has settled over the Land of Obfuscation.
“Even the Scientists Who Build AI Can’t Tell You How It Works” has a particularly interesting kicker:
“We built it, we trained it, but we don’t know what it’s doing.”
A group of artificial intelligence engineers struggling with the question, “What the heck is the system doing?” A click of the slide rule for MidJourney for this dramatic depiction of AI wizards at work.
The write up (which is an essay-interview confection) includes some thought-provoking comments. Here are three; you can visit the cited article for more scintillating insights:
Item 1: “… with reinforcement learning, you say, “All right, make this entire response more likely because the user liked it, and make this entire response less likely because the user didn’t like it.”
Item 2: “… The other big unknown that’s connected to this is we don’t know how to steer these things or control them in any reliable way. We can kind of nudge them
Item 3: “We don’t have the concepts that map onto these neurons to really be able to say anything interesting about how they behave.”
Item 4: “… we can sort of take some clippers and clip it into that shape. But that doesn’t mean we understand anything about the biology of that tree.”
Item 5: “… because there’s so much we don’t know about these systems, I imagine the spectrum of positive and negative possibilities is pretty wide.”
For more of this type of “explanation,” please, consult the source document cited above.
Several observations:
- I like the nudge and watch approach. Humanoids learning about what their code does may be useful.
- The nudging is subjective (human skill) and the reference to growing a tree and not knowing how that works exactly. Just do the bonsai thing. Interesting but is it efficient? Will it work? Sure or at least as Silicon Valley thinking permits
- The wide spectrum of good and bad. My reaction is to ask the striking writers and actors what their views of the bad side of the deal is. What if the writers get frisky and start throwing spit balls or (heaven forbid) old IBM Selectric type balls. Scary.
Net net: Perhaps Google knows best? Tensors, big computers, need for money, and control of advertising — I think I know why Google tries so hard to frame the AI discussion. A useful exercise is to compare what Google’s winner in the smart software power struggle has to say about Google’s vision. You can find that PR emission at this link. Be aware that the interviewer’s questions are almost as long at the interview subject’s answers. Does either suggest downsides comparable to the five items cited in this blog post?
Stephen E Arnold, July 18, 2023
Financial Analysts, Lawyers, and Consultants Can See Their Future
July 17, 2023
It is the middle of July 2023, and I think it is time for financial analysts, lawyers, and consultants to spruce up their résumés. Why would a dinobaby make such a suggestion to millions of the beloved Millennials, GenXers, the adorable GenY folk, and the vibrant GenZ lovers of TikTok, BMWs, and neutral colors?
I read three stories helpfully displayed by my trusty news reader. Let’s take a quick look at each and offer a handful of observations.
The first article is “This CEO Replaced 90% of Support Staff with an AI Chatbot.” The write up reports:
The chief executive of an Indian startup laid off 90% of his support staff after the firm built a chatbot powered by artificial intelligence that he says can handle customer queries much faster than his employees.
Yep, better, faster, and cheaper. Pick all three which is exactly what some senior managers will do. AI is now disrupting. But what about “higher skill” jobs than talking on the phone and looking up information for a clueless caller?
The second article is newsy or is it newsie? “Open AI and Associated Press Announce Partnership to Train AI on New Articles” reports:
[The deal] will see OpenAI licensing text content from the AP archives that will be used for training large language models (LLMs). In exchange, the AP will make use of OpenAI’s expertise and technology — though the media company clearly emphasized in a release that it is not using generative AI to help write actual news stories.
Will these stories become the property of the AP? Does Elon Musk have confidence in himself?
Young professionals learning that they are able to find their future elsewhere. In the MidJourney confection is a lawyer, a screenwriter, and a consultant at a blue chip outfit selling MBAs at five times the cost of their final year at university.
I think that the move puts Google in a bit of a spot if it processes AP content and a legal eagle can find that content in a Bard output. More significantly, hasta la vista reporters. Now the elimination of hard working, professional journalists will not happen immediately. However, from my vantage point in rural Kentucky, I hear the train a-rollin’ down the tracks. Whooo Whooo.
The third item is “Producers Allegedly Sought Rights to Replicate Extras Using AI, Forever, for Just $200.” The write up reports:
Hollywood’s top labor union for media professionals has alleged that studios want to pay extras around $200 for the rights to use their likenesses in AI – forever – for just $200.
Will the unions representing these skilled professionals refuse to cooperate? Does Elon Musk like Grimes’s music?
A certain blue chip consulting firm has made noises about betting $2 billion on smart software and Microsoft consulting. Oh, oh. Junior MBAs, it may not be too late to get an associate of arts degree in modern poetry so you can work as a prompt engineer. As a famous podcasting person says, “What say you?”
Several questions:
- Will trusted, reliable, research supporting real news organizations embrace smart software and say farewell to expensive humanoids?
- Will those making videos use computer generated entities?
- Will blue chip consulting firms find a way to boost partners’ bonuses standing on the digital shoulders of good enough software?
I sure hope you answered “no” to each of these questions. I have a nice two cruzeiro collectable from Brazil, circa 1952 to sell you. Make me an offer. Collectible currency is an alternative to writing prompts or becoming a tour guide in Astana. Oh, that’s in Kazakhstan.
Smart software is a cost reducer because humanoids [a] require salaries and health care, [b] take vacations, [c] create security vulnerabilities or are security vulnerabilities, and [d] require more than high school science club management methods related to sensitive issues.
Money and good enough will bring changes in news, Hollywood, and professional services.
Stephen E Arnold, July 17, 2023
-
- Subscribe to Beyond Search
Feature archive
News archive
-
