KONVERSATIONSMODELL. “När vi utvecklar dessa teknologier, pressar vi verkligen för allmän intelligens, allmänna förmågor med hög tillförlitlighet” menar Mira Murati i utvecklingen av ChatGPT och dess potentiella påverkan i framtiden.

Mira Murati är Chief Technology Officer på OpenAI och leder teamet bakom DALL-E och ChatGPT. DALL-E skapar konstverk med hjälp av AI, medan ChatGPT är en chatbot som kan svara på komplexa frågor på ett mänskligt sätt. Trots att Murati diskuterar de potentiella farorna med AI, lyckas Murati förmedla optimism om dess inverkan.

— Som med andra revolutioner som vi har gått igenom, kommer det att finnas nya jobb och vissa jobb kommer att gå förlorade, sa Murati till Trevor Noah i höstas om effekterna av AI, men jag är optimistisk.

ChatGPT blev en populär sökterm på Google och överträffade Bitcoin i januari. Även om vissa skolor förbjuder det, investerade Microsoft 10 miljarder dollar i OpenAI, och Google utfärdade en “kod röd”, oroad över att tekniken skulle kunna påverka dess sökverksamhet. ChatGPT:s outnyttjade potential och dess största svagheter diskuteras i en intervju med Mira Murati i Time Magazine. Murati berättar också om behovet av att reglera AI och vikten av att involvera olika perspektiv i samtalet.

ChatGPT har nyligen fått ett godkänt betyg på Wharton Business School MBA-provet och fyra stora juridikprov på universitetsnivå och Murati ser fram emot vad ChatGPT kan åstadkomma i framtiden.

Intervjuaren bad ChatGPT att ställa en fråga till Murati. Chatboten undrade då om begränsningarna eller utmaningarna hon stött på medans hon arbetat med ChatGPT och hur hon har övervunnit dem.

– Det är en bra fråga. ChatGPT är i grunden en stor konversationsmodell, ett stort neuralt nät som har tränats för att förutsäga nästa ord. Och utmaningarna med det är liknande utmaningar som vi ser med de grundläggande stora språkmodellerna: de kan utgöra fakta.

ChatGPTs självsäkra sätt att utgöra fakta är en av huvudutmaningarna.

– Ja. Detta är faktiskt en huvudutmaning. Vi valde dialoger specifikt eftersom dialoger är ett sätt att interagera med en modell och ge den feedback. Om vi ​​tror att svaret på modellen är felaktigt kan vi säga: “Är du säker? Jag tror faktiskt…” Och då har modellen en möjlighet att diskutera fram och tillbaka med dig, liknande hur vi skulle konversera med en annan människa.

Läs fortsättningen av intervjun med Murati HÄR.

The post Skaparen av ChatGPT anser att AI bör regleras first appeared on Opulens.

FUTURE. How can you build a picture of the future that isn’t dependent on the past? First ask yourself, does the way I see myself correlate or interfere with that future?

What’s your picture of the future?

This question leads to several other ones.

How do you envision it?

What is your place in it?

Are you the same person or do you differ from your current self somehow?

What are you feeling in this future?

Does it align with your dream life or are they two separate entities?

You may think of these questions as irrelevant and not related to your current situation. However, without answers to them, you can’t move forward and expect a certain result. As Einstein said, doing the same thing over and over again and expecting a different result, is called insanity.

You may have tried many times to lead a healthy lifestyle or change a defective habit but your attempts didn’t go far. It’s possible that you truly never saw your healthy self in that future, doing what’s needed. You could have kept trying the same methods without regard to different possibilities because it simply never occurred to you that you are capable of doing it differently. And that’s when your default future was created.  Don’t let the past define the future, use it as a learning experience, not as a measure of what’s to come.

When I realized that no matter how I picture my future, if I don’t change the way I see myself in it, I will keep crashing into my default view of it while my dreams will be somewhere else on a side. Once your view of yourself changes to reflect many possibilities of what can be, you can start changing your default future and your actions will align with it.

When our perception shifts, behavior follows. As I was creating a new vision of myself, I started doing what was needed to go along with that vision. Things started happening almost automatically, new circumstances began to unfold, new people and opportunities came about. Our minds miraculously create the reality we truly believe in.

How can you build a picture of the future that isn’t dependent on the past? First ask yourself, does the way I see myself correlate or interfere with that future?

If there are more inconsistencies in the way you see yourself now and how that self can create the desired future, things need to change in order for correlation to occur.

The post What’s your picture of the future? first appeared on Opulens.

IDEAS. Has it ever happened to you that you came across a product, an invention, or service and realized that it was your idea? How did it make you feel? As we know, only a small % of ideas turn into something that works.

Why does it happen? What is stopping people with great ideas? There can be a countless number of reasons why we stop before starting.

Firstly, it’s difficult. We are afraid to start something because it’s scary and unknown. For a long time, I was fully involved in someone else’s cause that wasn’t aligned with who I am just because I didn’t believe it was possible to develop my own.

At some point, I realized that it will be more difficult to keep investing myself in something I don’t believe in than to go on my own path. I may not see the whole road ahead, but the path in front is enough to keep moving forward.

Another typical reason is – it’s not a good time right now. We may have a million excuses for not starting something, and many of them may look valid at the time because in that state of mind we aren’t open to options and possibilities.

We find reasons for not starting now due to doubts that kill our motivation. Afterwards, we suffer and regret the lost time. I just started finding the antidote for my doubts, convincing myself that – Now is the time.

Analysis – paralysis. When we are unsure in which direction to move, we tend to overthink even things we were certain about. We tell ourselves that we don’t know enough, aren’t experienced enough, aren’t ready yet, still need to do more research.

At some point, I realized that I just need to act, otherwise everything will stay in the theory stage.

Last but not least, it can be about being stuck in a bad experience. When we dwell on something that didn’t work, we deprive ourselves of the lessons that we were supposed to learn, lessons that would aid us in the process further. This creates an unwillingness to continue trying.

It was essential for me to shift my mindset from beating myself up for bad decisions to taking them as learning experiences. Results are a reflection of our actions.

The post Might-Have-Been first appeared on Opulens.