Marco Camisani Calzolari - Professor of Digital Languages, IULM University - CEO @ Speakage & Founder DigitalGround UK

Articoli

• 338 – These Aren’t Fake News Anymore. This Is War on Science

  12 giugno, di Team — Artificial Decisions
  These Aren't Fake News Anymore. This Is War on Science Fake news was the easy version. Now there are people who take apart a doctor's reputation, deliberately, at a desk. And the favorite targets are science and medicine. Before, you made up a hoax and threw it online. That's not how it works now. It starts with a person. A name, a face. A whole campaign built to make them look incompetent, bought, dangerous. The person who treats you. Who explains how a vaccine works. Who tells you one drug works and another doesn't. Medicine and science are where it hurts most. Because there the distrust doesn't stay in idle talk. It becomes someone who quits their treatment. Someone who trusts the supplement bought online instead of the doctor who examined them. Here in the US scientists have stopped staying quiet. They're organizing, they've started to answer back, to defend themselves in public. In Europe the work is more on the rules. Two different roads, same problem underneath. And this is where AI comes in, and changes the scale of everything. A campaign that used to take time, money, people, you now set up in a few hours. Believable profiles, comments that look real, a thousand voices repeating the same accusation against the same name. Faster. More believable. Harder to take apart. For years we thought being right was enough. That competence defended itself. It doesn't. We're no longer defending the truth from a lie. We're defending competent people from those who want to pass them off as quacks. That's exactly the job of people who know how to communicate. And we left it uncovered. What do you think? #ArtificialDecisions #MCC #ArtificialIntelligence #Disinformation #Science 

• 337

  11 giugno, di Team — Artificial Decisions
  A Drug Used to Cost Ten Years and Two Billion. But AI Is Changing the Math A new drug, from the first idea to the day you find it on a pharmacy shelf, used to take ten years and over two billion euros. We take it for granted. That's how it's always worked. Now seventy-five Italian scientific societies, the ones representing hospital and university doctors, line up two numbers. Artificial intelligence cuts the cost of developing a drug by twenty-five percent. As some of you know, I hold a chair in cyberhumanities inside the medicine and surgery program at San Raffaele in Milan. Living here in New York I teach most of it remotely, but the upside is that I also get to work with the scientific community here in the US. And here in the United States the labs working with AI don't use it to replace researchers. They use it to read millions of molecules in a few days, throw out the ones that don't work and predict which ones are worth actually testing. The work that used to be years of trial and error now starts already filtered. But those ten years weren't all wasted. Part of it is research, and that AI speeds up. Another part is the clinical trials, real people you give the drug to see if it's safe. And that part is there to protect us. If we fall in love with faster and cheaper and cut there too, the risk is paid by whoever takes the pill. And if developing a drug costs twenty-five percent less, does that drug reach us any cheaper? We just have to remember that speed is good only as long as we don't pay for it with safety, and as long as we see that saving too. What do you think? #ArtificialDecisions #MCC #ArtificialIntelligence #DrugDiscovery #Healthcare 

• 335 – How to Fix Shadow AI, the AI Your Company Doesn’t Know Is There

  9 giugno, di Team — Artificial Decisions
  How to Fix Shadow AI, the AI Your Company Doesn't Know Is There Shadow AI is Artificial Intelligence running inside companies without anyone knowing. I talked about it before and it has gotten worse. I'll drop the link to that video in the comments. Here in the US every employee opens three or four AI tools a day, none of them approved. Inside go contracts, client emails, spreadsheets, source code. ChatGPT, Claude, Gemini… An employee clicks "sign in with Google" on a new tool, and that tool gets read and write access to the entire company Workspace. Drive, mail, documents. The traffic never touches the corporate network, so nobody sees it. Banning doesn't work. People will use AI anyway. 5 practical moves. First, a real inventory. OAuth audit, browser extension scan, internal survey. Ask "help us keep you safer" and they answer. Ask "are you breaking the rules" and they lie… Second, a policy written for humans. Approved tools listed, data that never goes into an AI, verified training opt-out for every tool, fast process to request new ones. Third, a fast lane. An employee who waits six weeks finds a workaround in two days. Fourth, browser-based monitoring, because that's where AI lives now. Personal devices too, but there only culture works… Fifth, just-in-time coaching. A short alert when someone opens an off-list tool, with the alternative ready. Thirty seconds worth more than a once-a-year training. Employees are the first entry point of attacks and have to become the first line of defense. Attention after a course or an incident fades in weeks. Without ongoing training, everything goes back to how it was. What do you think? #ArtificialDecisions #MCC #ArtificialIntelligence #ShadowAI #Cybersecurity 

• 334 – Nails the Interview. Can’t Do the Job

  8 giugno, di Team — Artificial Decisions
  Nails the Interview. Can't Do the Job A recruiter friend of mine here in New York told me she hired a perfect candidate. Clean answers, confident, sharp. First day on the job, he could barely do a thing. There's a program called Parakeet. It listens to your interview through the mic, reads the question, and writes the answer on your screen while the other person is still talking. You read it out loud. It's built to stay hidden: invisible on screen share in Zoom, Meet, Teams. It handles coding tests too, writing the code live. Fifty-two languages. About thirty dollars for a handful of interviews. So in an interview you're not judging the person anymore. You're judging how well they read what a machine feeds them. And plenty of people do it. Out of nearly twenty thousand interviews looked at between last summer and January, over a third got flagged as suspected cheating. Companies noticed. Google and McKinsey brought back in-person interviews. Seven hiring leaders out of ten here in the U.S. went back to meeting people face to face, for exactly this reason. The problem isn't using AI to prepare. That's studying, and it's fine. The problem is that people get into companies who passed the test but don't have the skills to be there. And the coworkers find out later, when they have to cover the gap. Unfortunately. AI is a great gym for walking into an interview ready. Used to read answers live, it prepares no one, and it fills workplaces with people who can't do the work. What do you think? #ArtificialDecisions #MCC #ArtificialIntelligence #Work #Hiring 

• 332 – Give Me a Face Like the AI Filters!

  6 giugno, di Team — Artificial Decisions
  Give Me a Face Like the AI Filters! A plastic surgeon I know here in New York told me more and more patients walk in with a photo in hand and ask him to rebuild their face exactly like that. And it's not a photo of them when they were younger. It's a face generated by Artificial Intelligence! Going to a surgeon to look younger, we've always done that, nothing new. But now they want features no human being has ever had: perfect symmetry, skin without a single pore, cheekbones drawn to the millimeter. A machine face. Because a filter has no bones, no skin that ages, no age at all. You have your cheekbone, it draws whatever cheekbone it likes. It spits out a smooth face, identical left and right, perfect the way only a machine can be. And the brain takes it for a goal within reach. We already saw this with Instagram and Snapchat filters. Ten-year-old girls, made up and dressed like adults, in front of the mirror hunting for a flaw that wasn't there before. Now it gets worse: the AI hands you back another species, a you that was never human. Now the app filters, toys really, get used as a mirror. A mirror that convinces us the wrong face is the real one. So, maybe AI will find the cure for all our fears. But meanwhile, in 2026, people walk into a surgeon's office holding the photo of a person who doesn't exist, and ask to become that thing! What do you think? #ArtificialDecisions #MCC #ArtificialIntelligence #AIFilters #DigitalBeauty 

0 | 5 | 10 | 15 | 20 | 25 | 30 | 35 | 40